From fdf60301f2585dd497d239d6fb17b533ce9a0334 Mon Sep 17 00:00:00 2001 From: Yichen Jiang Date: Mon, 24 Aug 2026 13:38:34 +0800 Subject: [PATCH 1/2] fix(system-prompt): stabilize workflow section order --- ...-20-config-hot-reload-resilience.i18n.yaml | 4 +-- ...2026-07-20-config-hot-reload-resilience.md | 2 +- ...6-07-20-config-hot-reload-resilience.zh.md | 2 +- ...system-prompt-section-order-ties.i18n.yaml | 6 ++++ ...-08-24-system-prompt-section-order-ties.md | 27 +++++++++++++++ ...-24-system-prompt-section-order-ties.zh.md | 27 +++++++++++++++ .../advanced-toolchain/session.1.jsonl | 2 +- .../advanced-toolchain/session.2.jsonl | 2 +- .../system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../both-mode-turn/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../code-mode-turn/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../lsp-definition/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../pty-tools/system-prompt.expected.md | 4 +-- .../read-image/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../system-prompt.1.expected.md | 4 +-- .../system-prompt.1.expected.md | 4 +-- .../session.1.jsonl | 2 +- .../session.2.jsonl | 2 +- .../system-prompt.1.expected.md | 4 +-- .../snapshots/subagent-mixed/session.1.jsonl | 2 +- .../snapshots/subagent-mixed/session.2.jsonl | 2 +- .../snapshots/subagent-multi/session.1.jsonl | 2 +- .../snapshots/subagent-multi/session.2.jsonl | 2 +- .../subagent-parallel/session.1.jsonl | 8 ++--- .../subagent-parallel/session.2.jsonl | 8 ++--- .../system-prompt.1.expected.md | 4 +-- .../text-turn/system-prompt.expected.md | 4 +-- .../web-fetch/system-prompt.expected.md | 4 +-- .../advanced-toolchain/session.1.jsonl | 10 +++--- .../advanced-toolchain/session.2.jsonl | 10 +++--- .../advanced-toolchain/session.jsonl | 28 +++++++-------- .../compaction-recovery/session.jsonl | 14 ++++---- .../tests/snapshots/pty-tools/session.jsonl | 34 +++++++++---------- packages/workflow/tool-ralph/src/index.ts | 2 +- packages/workflow/tool-workflow/src/index.ts | 2 +- 40 files changed, 164 insertions(+), 104 deletions(-) create mode 100644 .agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml create mode 100644 .agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md create mode 100644 .agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md diff --git a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.i18n.yaml b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.i18n.yaml index 0d29db621d..0b59ec54b6 100644 --- a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.i18n.yaml +++ b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.i18n.yaml @@ -2,5 +2,5 @@ # side as of the last confirmed-consistent state. Both languages carry equal authority; # after editing either side, bring the other along and re-record with: # pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.md -2026-07-20-config-hot-reload-resilience.md: 2b98ee73e1c7734da9c0ae6df5e3de1af6a5cbeb -2026-07-20-config-hot-reload-resilience.zh.md: 0a3652a7cac589b601d7fd30d0c713d4585be830 +2026-07-20-config-hot-reload-resilience.md: 6f8503c93bcbf55b1bd0543cf49e41cfe32fbe2f +2026-07-20-config-hot-reload-resilience.zh.md: 359a34a2252b2999f896245a4bd07f7da32d98f5 diff --git a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.md b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.md index 2b98ee73e1..6f8503c93b 100644 --- a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.md +++ b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.md @@ -38,4 +38,4 @@ HMR contains live refresh rejection. Its `registerConfig(filename, refresh)` met ## Testing -`packages/boot/app-boot/tests/config-reload.spec.ts` boots real temporary Loader/Include trees and covers parse and shape rejection, import-before-dispose, plugin/config restoration, multi-entry rollback, ancestor disablement, overlay convergence, option identity, failed direct-update persistence, and failed programmatic moves. `packages/boot/app-boot/tests/hmr-config.spec.ts` covers existing and missing exact paths, add/change/removal, serialized coalescing, disposal drainage, non-`Error` normalization, failure broadcast, and rejecting-observer containment. `packages/host/webserver/tests/webserver.spec.ts` proves a service-gated startup failure rejects Loader composition with its bind diagnostic, `packages/typert/loader/tests/loader.spec.ts` exercises awaited programmatic removal through a real Loader consumer, and the ACP `pty-tools` snapshot guards concurrent composition from reordering equal-priority prompt sections. +`packages/boot/app-boot/tests/config-reload.spec.ts` boots real temporary Loader/Include trees and covers parse and shape rejection, import-before-dispose, plugin/config restoration, multi-entry rollback, ancestor disablement, overlay convergence, option identity, failed direct-update persistence, and failed programmatic moves. `packages/boot/app-boot/tests/hmr-config.spec.ts` covers existing and missing exact paths, add/change/removal, serialized coalescing, disposal drainage, non-`Error` normalization, failure broadcast, and rejecting-observer containment. `packages/host/webserver/tests/webserver.spec.ts` proves a service-gated startup failure rejects Loader composition with its bind diagnostic, and `packages/typert/loader/tests/loader.spec.ts` exercises awaited programmatic removal through a real Loader consumer. diff --git a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.zh.md b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.zh.md index 0a3652a7ca..359a34a225 100644 --- a/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.zh.md +++ b/.agents/notes/implemented/bug-fix/2026-07-20-config-hot-reload-resilience.zh.md @@ -38,4 +38,4 @@ HMR 收容实时刷新 rejection。其 `registerConfig(filename, refresh)` 方 ## Testing -`packages/boot/app-boot/tests/config-reload.spec.ts` 启动真实的临时 Loader/Include 树，并覆盖对解析和形状错误的拒绝、先导入再 dispose、插件/配置恢复、多配置项回滚、祖先禁用、overlay 收敛、option 对象身份、失败的直接更新不持久化以及失败的程序化移动。`packages/boot/app-boot/tests/hmr-config.spec.ts` 覆盖现有和缺失的确切路径、添加/变更/移除、串行化合并、dispose 排空、非 `Error` 值的规范化、失败广播以及对发生 rejection 的观察者的收容。`packages/host/webserver/tests/webserver.spec.ts` 证明受服务门控的启动失败会让 Loader 组合以其 bind 诊断 reject；`packages/typert/loader/tests/loader.spec.ts` 则通过真实 Loader 消费方演练可等待的程序化移除；ACP（Agent Client Protocol）的 `pty-tools` 快照会防止并发组合改变同优先级提示词段的顺序。 +`packages/boot/app-boot/tests/config-reload.spec.ts` 启动真实的临时 Loader/Include 树，并覆盖对解析和形状错误的拒绝、先导入再 dispose、插件/配置恢复、多配置项回滚、祖先禁用、overlay 收敛、option 对象身份、失败的直接更新不持久化以及失败的程序化移动。`packages/boot/app-boot/tests/hmr-config.spec.ts` 覆盖现有和缺失的确切路径、添加/变更/移除、串行化合并、dispose 排空、非 `Error` 值的规范化、失败广播以及对发生 rejection 的观察者的收容。`packages/host/webserver/tests/webserver.spec.ts` 证明受服务门控的启动失败会让 Loader 组合以其 bind 诊断 reject；`packages/typert/loader/tests/loader.spec.ts` 则通过真实 Loader 消费方演练可等待的程序化移除。 diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml new file mode 100644 index 0000000000..d53f6820ce --- /dev/null +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml @@ -0,0 +1,6 @@ +# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each +# side as of the last confirmed-consistent state. Both languages carry equal authority; +# after editing either side, bring the other along and re-record with: +# pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md +2026-08-24-system-prompt-section-order-ties.md: 9aba5ee117a07b1d5abd6497232e61a48e55a639 +2026-08-24-system-prompt-section-order-ties.zh.md: 9736aa94f12a9ddf7242ef6d95b3c0204c4b62ed diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md new file mode 100644 index 0000000000..9aba5ee117 --- /dev/null +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md @@ -0,0 +1,27 @@ +# Agent Note: Equal-order system-prompt sections render in activation order + +Status: implemented + +English | [中文](2026-08-24-system-prompt-section-order-ties.zh.md) + +## Problem + +`SystemPromptRegistry` sorts sections by `order` with a stable sort, so equal orders render in plugin-activation order. `tool:cordis` and `tool:workflow` both declared `order: 115`, while their activation order varies between clean platform compositions. ACP and SDK snapshot replays could therefore assemble the same sections in a different order from their committed `system-prompt.expected.md` files. + +## Decision + +Give the three ordered sections distinct values: `tool:cordis` stays at 115, `tool:workflow` moves to 116, and `tool:ralph` moves from 116 to 117 so it remains after workflow. Prompt text and tool schemas remain unchanged. + +## Alternatives considered + +**Normalize section order in the snapshot harness.** Rejected because the runtime, request header, and model prompt would remain sensitive to activation timing while only the fixture comparison hid the difference. + +**Tie-break equal orders by section name in the registry.** Rejected because it would silently reorder every existing tie. Explicit orders keep each model-visible placement local to the contributing plugin. + +## Consequences + +The Cordis, workflow, and Ralph guidance has one platform-independent order. Prompt-section placements that require a stable relative position need distinct `order` values; stable sorting continues to preserve activation order for intentional ties. + +## Testing + +The keyless ACP and SDK snapshot replays pin the affected system-prompt order, and the full snapshot suite verifies the refreshed fixtures. diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md new file mode 100644 index 0000000000..9736aa94f1 --- /dev/null +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md @@ -0,0 +1,27 @@ +# Agent Note: 等序系统提示词分段按激活顺序渲染 + +Status: implemented + +[English](2026-08-24-system-prompt-section-order-ties.md) | 中文 + +## Problem + +`SystemPromptRegistry` 使用稳定排序按 `order` 排列分段，因此相同 order 的分段会按插件激活顺序渲染。`tool:cordis` 与 `tool:workflow` 都声明了 `order: 115`，但两者在不同平台的全新组合中激活顺序不同。因此，ACP（Agent Client Protocol）与 SDK 的快照回放可能把相同分段组装成不同于已提交 `system-prompt.expected.md` 文件的顺序。 + +## Decision + +为这三个有序分段指定互不相同的 order：`tool:cordis` 保持 115，`tool:workflow` 移到 116，`tool:ralph` 从 116 移到 117，从而继续排在 workflow 之后。提示词文本与工具 schema 保持不变。 + +## Alternatives considered + +**在快照 harness 中规范化分段顺序。** 已否决，因为运行时、请求标头和模型提示词仍然受激活时序影响，只有 fixture 比较会隐藏差异。 + +**在注册表中用分段名称打破并列。** 已否决，因为这会静默重排每一组现有并列。显式 order 让每个模型可见位置都由贡献该分段的插件就地决定。 + +## Consequences + +Cordis、workflow 与 Ralph 指引具有不依赖平台的唯一顺序。需要稳定相对位置的提示词分段必须使用互不相同的 `order`；对于有意设置的并列，稳定排序仍会保留激活顺序。 + +## Testing + +无密钥 ACP 与 SDK 快照回放会固定受影响的系统提示词顺序，完整快照套件则验证刷新的 fixture。 diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl index 01ba355631..470217b161 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Check direct child"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebe0cfa0-a909-47e0-8294-28ad84a8fe77"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"e3c23441-606f-4e7a-8338-b434c0d04a4e"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"3d72b77d-f8f9-4e5b-b918-6ef0d13c35a2"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly DIRECT_CHILD_OK and","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl index 73958c517c..a62cd266ab 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"2ac2cc54-9bce-4cfa-a569-a64f51bc30a7"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"5d128e81-c7c2-4cd0-ad1c-7409b33650fc"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"2c50b91f-f7ee-4aad-afc5-b08732633604"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly WORKFLOW_CHILD_OK and","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md index 1743643d95..c20ff88a34 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md @@ -131,10 +131,10 @@ return { Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md index 7150bf2e6b..4517bb46cb 100644 --- a/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md index 37df6287ed..0c8262c90d 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md @@ -25,10 +25,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md index daf622df60..4913a69d8c 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md @@ -27,10 +27,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md index 6894f13fb6..79e1d537f8 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md @@ -27,10 +27,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md index 9b4698844c..95936c6f15 100644 --- a/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md @@ -18,6 +18,6 @@ Use the web_search tool to discover current information on the web. The required Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md index b906b6f3c8..fee88385ae 100644 --- a/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md index 545e903230..f37a2d570a 100644 --- a/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md index 06b614520c..9939935412 100644 --- a/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md index a0d3386eaa..dd0ebca77d 100644 --- a/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md index 800356dccc..5166d0236f 100644 --- a/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md index b198b48a12..2da38d57c5 100644 --- a/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md index b198b48a12..2da38d57c5 100644 --- a/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl index b64494a028..2e7f6f7c9a 100644 --- a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Start depth one"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Call subagent once. Ask that child to attempt one further subagent call, then report the result."}],"source":{"kind":"user"},"role":"user","id":"a8129357-1bde-4cbd-90b4-6b8ad51d52e1"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"72272791-eefd-48f8-94da-02b132ae9d2a"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"37b2a8d1-90f2-4aad-9e35-737559cdaaa2"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Call subagent once. Ask that","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl index 3e3dfa357b..3a3d440f2d 100644 --- a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Start depth two"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Attempt one subagent call beyond the configured cap, then report the rejection."}],"source":{"kind":"user"},"role":"user","id":"d4dc5a16-e542-4dd9-8e82-e6b7829cfc4b"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"c54120cc-6a7f-41f6-a71d-42b4805fa2ca"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"776d4567-6a92-45a6-8a74-8d376af180e5"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Attempt one subagent call beyond","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md index b198b48a12..2da38d57c5 100644 --- a/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl index b64a6cfdfe..474ccadf60 100644 --- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Reply ALPHA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"73ce401a-faaf-408a-879e-7485380d537d"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"f216ca0e-6dcc-4ab3-9cdb-fe38d3dacca2"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"c3bc9e3e-bf25-4169-ad2c-bda0183fad84"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl index e7de6e1e7d..8ff90306d3 100644 --- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl @@ -31,7 +31,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"fork","label":"Recall project codeword"}} {"type":"step/start","data":{"turn":2,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else."}],"source":{"kind":"user"},"role":"user","id":"86e9f144-764f-460d-b72b-262cffe43d77"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"ac4f4d97-639d-4ad0-a513-219a58355531"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"95dee7dd-6ac8-4695-b3f0-745cb724dfbd"},"surfaceOp":"append"} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}} {"type":"assistant/chunk","data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} {"type":"reasoning-chunks","data":{"turn":2,"step":1,"index":0,"dt":[0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"texts":["The"," user"," is"," asking"," me"," to"," recall"," the"," project"," cod","ew","ord"," that"," was"," mentioned"," earlier"," in"," the"," conversation","."," I"," was"," told"," to"," remember"," it",":"," SA","FF","RON","."]}} diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl index e1b186cd1f..6cea54fb30 100644 --- a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Return ALPHA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"a287f842-f6f2-4a17-ab4c-820e41f498d5"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"47cdc6a0-a8c8-4842-964a-ad4bc97dc76a"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"967aa8c4-7ad8-4973-966a-a6fcf8ba00dd"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl index f8cb2e5924..2888c8f591 100644 --- a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Return BETA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word BETA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"53f6419d-8ddc-4eee-8803-5b68411336f9"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"f9132345-93c9-40c0-b489-5916bbca96bc"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"c52ae99e-3f24-4803-a002-56564ed59845"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl index 14b644dc33..0cb0df4130 100644 --- a/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl @@ -2,19 +2,19 @@ {"type":"sandbox/mode","data":{"mode":"danger-full-access","source":"delegation"}} {"type":"approval/policy","data":{"policy":"never","source":"delegation"}} {"type":"permission/preset","data":{"preset":"danger-full-access"}} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"fadafbc9-263b-4169-82c6-a39868629377"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c0a22dc3-7ed8-401f-9d27-daa216035e79"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Say the word ALPHA"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"fadafbc9-263b-4169-82c6-a39868629377"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"1d6d2982-78f7-49b9-b32d-0eb465d672b1"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c0a22dc3-7ed8-401f-9d27-daa216035e79"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"f995fac5-e34b-468d-9162-d64cbac11ab8"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ALPHA"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"9ccb6b64-4dfb-47a2-9967-13ab05483998"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"2814fde1-ee68-4d3c-a14f-7b9ea936384c"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl index 5f298c76c5..86ae5204d3 100644 --- a/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl @@ -2,19 +2,19 @@ {"type":"sandbox/mode","data":{"mode":"danger-full-access","source":"delegation"}} {"type":"approval/policy","data":{"policy":"never","source":"delegation"}} {"type":"permission/preset","data":{"preset":"danger-full-access"}} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"dc34a17f-fb30-4afe-a11f-a0d8a1d51658"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"6bbb164f-99ee-4b7f-a0bd-c1a35b02a076"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Say the word ALPHA"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"dc34a17f-fb30-4afe-a11f-a0d8a1d51658"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"12a26f3d-f11e-4de4-8bed-d997590d21e0"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"6bbb164f-99ee-4b7f-a0bd-c1a35b02a076"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"7cbd146e-97ac-49f2-979a-4221d9277ec8"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ALPHA"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d368f9a5-7d0a-46f0-a7d8-10e1fafa1e74"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"c40ec82f-7cb8-4885-be04-ba85bff6e2a2"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md index b198b48a12..2da38d57c5 100644 --- a/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md index 975b5a7baf..b1d7332947 100644 --- a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md index b70cc036d4..ab93903173 100644 --- a/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. - Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + +Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl index 1287de6339..b1e3b0e457 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl @@ -1,19 +1,19 @@ {"type":"session","version":0,"id":"22222222-2222-4222-8222-222222222222","createdAt":1783950001000,"cwd":"{{cwd}}","parentSession":"11111111-1111-4111-8111-111111111111","origin":"subagent","delegationDepth":1} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c66e310e-2597-4d01-85c8-2d70a9d831c0"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"f9b76611-65ac-4d61-ba91-cb003d56670b"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Check direct child"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c66e310e-2597-4d01-85c8-2d70a9d831c0"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"fd0a0587-8df4-46b2-809a-317346a4c0f4"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"f9b76611-65ac-4d61-ba91-cb003d56670b"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"09b80e2e-0a74-48b9-88c3-03e1d859fb5d"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly DIRECT_CHILD_OK and","messageSeqs":[5],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. ONLY what you print or return comes back to you — intermediate tool results never enter the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return comes back — curate it.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"DIRECT_CHILD_OK"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DIRECT_CHILD_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"02dd8a61-a39a-46d0-8f6f-457533271cae"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"f018f8fa-7a81-46a3-93a0-a30705a7785f"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl index 1f18e866a9..6be8abb48f 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl @@ -1,19 +1,19 @@ {"type":"session","version":0,"id":"33333333-3333-4333-8333-333333333333","createdAt":1783950002000,"cwd":"{{cwd}}","parentSession":"11111111-1111-4111-8111-111111111111","origin":"subagent","delegationDepth":1} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"8b3cd23c-82f1-4903-8a3f-b9082059b40c"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebb3764d-0742-4054-a421-69d479b90021"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"8b3cd23c-82f1-4903-8a3f-b9082059b40c"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"823e5037-9e96-4ef5-8c5b-cbe73b993ee2"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebb3764d-0742-4054-a421-69d479b90021"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"8390dc9d-26c3-439b-9f60-ac263964f0d3"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly WORKFLOW_CHILD_OK and","messageSeqs":[5],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. ONLY what you print or return comes back to you — intermediate tool results never enter the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return comes back — curate it.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"WORKFLOW_CHILD_OK"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"WORKFLOW_CHILD_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"cd78f077-1fad-4cdc-ab56-09d39d9095cd"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"80f282de-e531-41cc-94d9-d6896f6a2800"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl index c49d9b2745..dbe2d47352 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl @@ -1,20 +1,20 @@ {"type":"session","version":0,"id":"11111111-1111-4111-8111-111111111111","createdAt":1783950000000,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"8a0ac233-283c-4eb4-8bbd-5c50b7e99afe"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"be704496-d340-4e19-8506-6cf81feb739a"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"8a0ac233-283c-4eb4-8bbd-5c50b7e99afe"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"be704496-d340-4e19-8506-6cf81feb739a"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Run this advanced flow exactly","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. ONLY what you print or return comes back to you — intermediate tool results never enter the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return comes back — curate it.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"advanced-define","name":"cordis_define","argumentsDelta":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e0351488-7bca-48d6-b7af-87d533858f47"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"28ed15b0-0b10-4cb7-bdb8-17af0c101f9a"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"advanced-define"},"content":[{"type":"tool-result","toolCallId":"advanced-define","content":[{"type":"text","text":"Defined snap-1/pkg-1 (Snapshot Marker); it is not running yet. Use cordis_run to activate this Package."}],"isError":false}],"role":"user","id":"9fd9ba62-1b6d-4bb7-98c6-97815e983026"},"meta":{"pluginId":"snap-1","packageId":"pkg-1"}},"sourceEventSeqs":[14],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"advanced-define"},"content":[{"type":"tool-result","toolCallId":"advanced-define","content":[{"type":"text","text":"Defined snap-1/pkg-1 (Snapshot Marker); it is not running yet. Use cordis_run to activate this Package."}],"isError":false}],"role":"user","id":"f037d5e8-3ef3-4c76-8bd9-e823543d16c2"},"meta":{"pluginId":"snap-1","packageId":"pkg-1"}},"sourceEventSeqs":[14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -22,13 +22,13 @@ {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"6f6c3cc3-350b-4afb-a3d1-8f91b9494628"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[18,19,20,21,22],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"8a02e55c-d6e7-4b63-9674-856c71c385a1"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[18,19,20,21,22],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":2,"callId":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}} {"type":"tool/code-dispatch-start","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_run","arguments":{"pluginId":"snap-1","packageId":"pkg-1","mode":"run"}}} {"type":"tool/code-dispatch","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_run","arguments":{"pluginId":"snap-1","packageId":"pkg-1","mode":"run"},"isError":false,"content":[{"type":"text","text":"snap-1/pkg-1 is running (run-1)."}]}} {"type":"tool/code-dispatch-start","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:2","name":"cordis_inspect_self","arguments":{"pluginId":"snap-1"}}} {"type":"tool/code-dispatch","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:2","name":"cordis_inspect_self","arguments":{"pluginId":"snap-1"},"isError":false,"content":[{"type":"text","text":"{\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n}"}]}} -{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"advanced-code"},"content":[{"type":"tool-result","toolCallId":"advanced-code","content":[{"type":"text","text":"{\n \"run\": {\n \"status\": \"running\",\n \"pluginId\": \"snap-1\",\n \"packageId\": \"pkg-1\",\n \"pluginRunId\": \"run-1\",\n \"currentPackageId\": \"pkg-1\",\n \"host\": {\n \"status\": \"running\",\n \"provides\": [],\n \"waitingFor\": []\n },\n \"client\": {\n \"status\": \"absent\",\n \"waitingFor\": []\n }\n },\n \"inspected\": {\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n }\n}"}],"isError":false}],"role":"user","id":"40a17cbf-a853-4813-bbb5-7970cfbc7010"}},"sourceEventSeqs":[24],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"advanced-code"},"content":[{"type":"tool-result","toolCallId":"advanced-code","content":[{"type":"text","text":"{\n \"run\": {\n \"status\": \"running\",\n \"pluginId\": \"snap-1\",\n \"packageId\": \"pkg-1\",\n \"pluginRunId\": \"run-1\",\n \"currentPackageId\": \"pkg-1\",\n \"host\": {\n \"status\": \"running\",\n \"provides\": [],\n \"waitingFor\": []\n },\n \"client\": {\n \"status\": \"absent\",\n \"waitingFor\": []\n }\n },\n \"inspected\": {\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n }\n}"}],"isError":false}],"role":"user","id":"0d0af8e0-841c-416f-9be6-bcbed9406669"}},"sourceEventSeqs":[24],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"step/start","data":{"turn":1,"step":3}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -36,9 +36,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"8bdea090-4a9a-451f-bb21-459af50472fa"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e1775d38-b17e-4e98-8dff-700d604cddbf"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":3,"callId":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}} -{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"advanced-direct-child"},"content":[{"type":"tool-result","toolCallId":"advanced-direct-child","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"isError":false}],"role":"user","id":"c978c208-8ebd-4dd6-b997-606ca7de787e"}},"sourceEventSeqs":[38],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"advanced-direct-child"},"content":[{"type":"tool-result","toolCallId":"advanced-direct-child","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"isError":false}],"role":"user","id":"a9ae23a0-f3b1-432e-ade5-0a25aec04838"}},"sourceEventSeqs":[38],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":3}} {"type":"step/start","data":{"turn":1,"step":4}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -46,13 +46,13 @@ {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e55b2b2e-497c-45d2-8115-16f317ae573f"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"3be25390-4ab3-4d98-a7b3-6b7df9c89f47"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":4,"callId":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}} {"type":"tool-workflow/run-start","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","name":"advanced-headless-snapshot"}} {"type":"tool-workflow/agent-start","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","seq":1,"label":"workflow-child","phase":"Delegate","childId":"33333333-3333-4333-8333-333333333333"}} {"type":"tool-workflow/agent-end","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","seq":1,"outcome":"completed"}} {"type":"tool-workflow/run-end","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","stopReason":"completed"}} -{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"advanced-workflow"},"content":[{"type":"tool-result","toolCallId":"advanced-workflow","content":[{"type":"text","text":"workflow \"advanced-headless-snapshot\" completed (1 agent).\nReturn value:\n{\n \"reply\": \"WORKFLOW_CHILD_OK\"\n}"}],"isError":false}],"role":"user","id":"f85f58fc-8d7e-4c9f-a0fe-caff480a9fec"}},"sourceEventSeqs":[48],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"advanced-workflow"},"content":[{"type":"tool-result","toolCallId":"advanced-workflow","content":[{"type":"text","text":"workflow \"advanced-headless-snapshot\" completed (1 agent).\nReturn value:\n{\n \"reply\": \"WORKFLOW_CHILD_OK\"\n}"}],"isError":false}],"role":"user","id":"4d88debc-419d-4bea-ace5-e4626475cf83"}},"sourceEventSeqs":[48],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":4}} {"type":"step/start","data":{"turn":1,"step":5}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -60,9 +60,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"68bd993f-a0bd-4ea8-ad16-6b1a19e09bd3"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[56,57,58,59,60],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"00552a01-9bbc-4fc0-82b5-a00140d5b12a"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[56,57,58,59,60],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":5,"callId":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}} -{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"advanced-undefine"},"content":[{"type":"tool-result","toolCallId":"advanced-undefine","content":[{"type":"text","text":"Removed dynamic Plugin snap-1 and all of its Packages."}],"isError":false}],"role":"user","id":"2d90e3b5-2a4c-4408-a1a0-3d009786a07b"}},"sourceEventSeqs":[62],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"advanced-undefine"},"content":[{"type":"tool-result","toolCallId":"advanced-undefine","content":[{"type":"text","text":"Removed dynamic Plugin snap-1 and all of its Packages."}],"isError":false}],"role":"user","id":"af8f561e-1a9a-48e9-a94b-80e361fe5142"}},"sourceEventSeqs":[62],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":5}} {"type":"step/start","data":{"turn":1,"step":6}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} @@ -70,6 +70,6 @@ {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ADVANCED_HEADLESS_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"text","text":"ADVANCED_HEADLESS_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"f3823367-2e25-43d2-a129-70dc492b2a90"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[66,67,68,69,70],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"text","text":"ADVANCED_HEADLESS_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4556caeb-f947-40b9-a3c4-98e0f48b2edc"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[66,67,68,69,70],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":6}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl b/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl index 575bbc5dfa..5963280e2c 100644 --- a/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl +++ b/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl @@ -1,32 +1,32 @@ {"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"10eb2388-2d40-4564-af27-e7a5419fc14e"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"7e4059c6-45e3-4067-8966-b2d402b97879"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"10eb2388-2d40-4564-af27-e7a5419fc14e"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"7e4059c6-45e3-4067-8966-b2d402b97879"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Establish a durable compaction premise","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash","contextWindow":128000}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_compaction_marker","name":"bash","argumentsDelta":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":24,"outputTokens":6}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"a71b2cfd-c18f-4a1b-82f6-e89fb371a87e"},"usage":{"inputTokens":24,"outputTokens":6}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d1e07317-e842-4a27-bd7d-b679adebb42a"},"usage":{"inputTokens":24,"outputTokens":6}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_compaction_marker"},"content":[{"type":"tool-result","toolCallId":"call_compaction_marker","content":[{"type":"text","text":"alpha\n"}],"isError":false}],"role":"user","id":"c9e68608-2dff-44bc-a344-b01006272378"}},"sourceEventSeqs":[14],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_compaction_marker"},"content":[{"type":"tool-result","toolCallId":"call_compaction_marker","content":[{"type":"text","text":"alpha\n"}],"isError":false}],"role":"user","id":"ed26c0aa-4a0a-49e3-88b4-795f48e02e45"}},"sourceEventSeqs":[14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"error","failure":{"message":"snapshot request exceeded the model context window","code":"CONTEXT_WINDOW_EXCEEDED"}}}}} {"type":"compaction/start","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","turn":1}} {"type":"compaction/summary","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","summary":[{"type":"text","text":"The request established a durable compaction premise."}],"rawOutput":[{"type":"text","text":"The request established a durable compaction premise."}],"llmStreamCall":true,"shadowedRange":{"start":4,"end":4},"shadowedSeqs":[4],"shadowedTokenCount":266,"provider":"deepseek-official","model":"deepseek-v4-flash","maxTokens":32,"usage":{"inputTokens":20,"outputTokens":4}}} -{"type":"user/message","data":{"content":[{"type":"text","text":"This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.\n\n"},{"type":"text","text":"The request established a durable compaction premise."},{"type":"text","text":""}],"source":{"kind":"plugin","plugin":"compact","compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1"},"role":"user","id":"3668b957-07a2-4cb7-96b1-98a23ac8cdb8"},"sourceEventSeqs":[19,20,4],"surfaceOp":{"op":"replace","start":4,"end":4}} +{"type":"user/message","data":{"content":[{"type":"text","text":"This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.\n\n"},{"type":"text","text":"The request established a durable compaction premise."},{"type":"text","text":""}],"source":{"kind":"plugin","plugin":"compact","compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1"},"role":"user","id":"6f0985ae-7ff0-4923-80ab-0548d2cf466f"},"sourceEventSeqs":[19,20,4],"surfaceOp":{"op":"replace","start":4,"end":4}} {"type":"compaction/end","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","turn":1}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":0,"text":"COMPACTION RECOVERED"}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"COMPACTION RECOVERED"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":20,"outputTokens":4}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"text","text":"COMPACTION RECOVERED"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"bcbfd4ff-60e5-4634-ae39-4de3708a8abc"},"usage":{"inputTokens":20,"outputTokens":4}},"sourceEventSeqs":[23,24,25,26,27],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"text","text":"COMPACTION RECOVERED"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e5a0b0a1-8f90-450c-b994-27eb0d084be0"},"usage":{"inputTokens":20,"outputTokens":4}},"sourceEventSeqs":[23,24,25,26,27],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl b/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl index 6e5148b4de..b285d23511 100644 --- a/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl +++ b/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl @@ -1,21 +1,21 @@ {"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d35cdacd-b5e6-4968-b7a3-5ec48f403ef7"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d20ef69c-ac94-4aed-9c4a-821a6ebff160"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d35cdacd-b5e6-4968-b7a3-5ec48f403ef7"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}]},"role":"user","id":"053af702-9950-4860-913a-3c7e45a54f9d"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d20ef69c-ac94-4aed-9c4a-821a6ebff160"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}]},"role":"user","id":"8731a232-5605-410b-9dd4-f7ae7b1a3f04"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Exercise the six PTY tools","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nUse a terminal session only when work needs persistent terminal state or interactive stdin; prefer shell/read/write/edit for bounded one-shot operations. Track every terminal session id and close sessions that no longer matter. An inferred_idle or timeout result does not prove the foreground command exited.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"terminal_close","description":"Close one persistent terminal and wait until its captured owned process tree is gone.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."}},"required":["sessionId"]}},{"name":"terminal_list","description":"List persistent terminal sessions owned by the current agent.","parameters":{"type":"object","properties":{}}},{"name":"terminal_open","description":"Create a persistent, owner-isolated terminal session from a registered backend type. Use this for shell or REPL state that must survive across tool calls.","parameters":{"type":"object","properties":{"type":{"type":"string","description":"Registered terminal backend type, usually \"shell\"."},"name":{"type":"string","description":"Optional owner-local display name such as \"main\" or \"gdb\"."},"cwd":{"type":"string","description":"Initial working directory. Defaults to the deployment workspace root."}},"required":["type"]}},{"name":"terminal_read","description":"Read a bounded page of retained output from a persistent terminal without sending input.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"offset":{"type":"number","description":"Newest-relative line offset (default 0)."},"count":{"type":"number","description":"Requested line count (default 500; backend caps apply)."}},"required":["sessionId"]}},{"name":"terminal_send","description":"Send text to a persistent terminal. By default Enter is submitted and the call waits for a prompt, stdin wait, output silence, timeout, or session exit. Background mode returns a job id for job_output/job_kill.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id returned by terminal_open or terminal_list."},"text":{"type":"string","description":"UTF-8 text to write to the terminal."},"submit":{"type":"boolean","description":"Submit Enter after text (default true). Set false for control characters or incomplete REPL input."},"run_in_background":{"type":"boolean","description":"Return a job id immediately; collect with job_output or stop with job_kill."}},"required":["sessionId","text"]}},{"name":"terminal_signal","description":"Send an allowed signal to the current foreground process group of a persistent terminal.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"signal":{"type":"string","description":"Signal to deliver. Shell-targeted SIGKILL is rejected; use terminal_close.","enum":["SIGINT","SIGTERM","SIGKILL","SIGTSTP","SIGHUP"]}},"required":["sessionId","signal"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nUse a terminal session only when work needs persistent terminal state or interactive stdin; prefer shell/read/write/edit for bounded one-shot operations. Track every terminal session id and close sessions that no longer matter. An inferred_idle or timeout result does not prove the foreground command exited.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"terminal_close","description":"Close one persistent terminal and wait until its captured owned process tree is gone.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."}},"required":["sessionId"]}},{"name":"terminal_list","description":"List persistent terminal sessions owned by the current agent.","parameters":{"type":"object","properties":{}}},{"name":"terminal_open","description":"Create a persistent, owner-isolated terminal session from a registered backend type. Use this for shell or REPL state that must survive across tool calls.","parameters":{"type":"object","properties":{"type":{"type":"string","description":"Registered terminal backend type, usually \"shell\"."},"name":{"type":"string","description":"Optional owner-local display name such as \"main\" or \"gdb\"."},"cwd":{"type":"string","description":"Initial working directory. Defaults to the deployment workspace root."}},"required":["type"]}},{"name":"terminal_read","description":"Read a bounded page of retained output from a persistent terminal without sending input.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"offset":{"type":"number","description":"Newest-relative line offset (default 0)."},"count":{"type":"number","description":"Requested line count (default 500; backend caps apply)."}},"required":["sessionId"]}},{"name":"terminal_send","description":"Send text to a persistent terminal. By default Enter is submitted and the call waits for a prompt, stdin wait, output silence, timeout, or session exit. Background mode returns a job id for job_output/job_kill.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id returned by terminal_open or terminal_list."},"text":{"type":"string","description":"UTF-8 text to write to the terminal."},"submit":{"type":"boolean","description":"Submit Enter after text (default true). Set false for control characters or incomplete REPL input."},"run_in_background":{"type":"boolean","description":"Return a job id immediately; collect with job_output or stop with job_kill."}},"required":["sessionId","text"]}},{"name":"terminal_signal","description":"Send an allowed signal to the current foreground process group of a persistent terminal.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"signal":{"type":"string","description":"Signal to deliver. Shell-targeted SIGKILL is rejected; use terminal_close.","enum":["SIGINT","SIGTERM","SIGKILL","SIGTSTP","SIGHUP"]}},"required":["sessionId","signal"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"pty-spawn","name":"terminal_open","argumentsDelta":"{\"type\":\"shell\",\"name\":\"main\"}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"911213f8-acce-47be-a4f2-9d72ef55d83a"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d642809d-44ee-42cb-867d-63acb3d88b7a"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"pty-spawn"},"content":[{"type":"tool-result","toolCallId":"pty-spawn","content":[{"type":"text","text":"started terminal session pty-1 (main) [type: shell]\ndsh> "}],"isError":false}],"role":"user","id":"2da21b47-7fb3-444c-99a6-2c21743731ee"}},"sourceEventSeqs":[15],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"pty-spawn"},"content":[{"type":"tool-result","toolCallId":"pty-spawn","content":[{"type":"text","text":"started terminal session pty-1 (main) [type: shell]\ndsh> "}],"isError":false}],"role":"user","id":"f5d9b371-8400-4993-9cac-f61fcc1aab9b"}},"sourceEventSeqs":[15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -23,9 +23,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"cf12d7ee-322a-4057-97b0-98d828a96f1a"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"bf54676e-46ad-41b6-8938-c5b5cf9429ea"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":2,"callId":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}} -{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"pty-send"},"content":[{"type":"tool-result","toolCallId":"pty-send","content":[{"type":"text","text":"K\ndsh> \n[wait: stdin_read]\n[session: running]\n[output truncated]"}],"isError":false}],"role":"user","id":"d1ffb2dc-6a33-4c9e-aeb9-b87bf6da8617"},"meta":{"viewport":"printf 'PTY_OK\\n'\nPTY_OK\ndsh> ","waitReason":"stdin_read","sessionStatus":{"kind":"running"},"truncated":false}},"sourceEventSeqs":[25],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"pty-send"},"content":[{"type":"tool-result","toolCallId":"pty-send","content":[{"type":"text","text":"K\ndsh> \n[wait: stdin_read]\n[session: running]\n[output truncated]"}],"isError":false}],"role":"user","id":"1ea99fa1-54bd-45cd-8935-041298013a91"},"meta":{"viewport":"printf 'PTY_OK\\n'\nPTY_OK\ndsh> ","waitReason":"stdin_read","sessionStatus":{"kind":"running"},"truncated":false}},"sourceEventSeqs":[25],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"step/start","data":{"turn":1,"step":3}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -33,9 +33,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b9b99ae9-4685-4ee5-b951-fbe58f84c4e3"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[29,30,31,32,33],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"cb394133-a39a-4c7c-bba0-f51a06a1ddd9"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[29,30,31,32,33],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":3,"callId":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}} -{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"pty-read"},"content":[{"type":"tool-result","toolCallId":"pty-read","content":[{"type":"text","text":"dsh> printf 'PTY_OK\\n'\nPTY_OK\ndsh> \n[lines: 0-3 of 3]"}],"isError":false}],"role":"user","id":"d0d7331d-0d54-457a-9f7b-beb22abd34e6"}},"sourceEventSeqs":[35],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"pty-read"},"content":[{"type":"tool-result","toolCallId":"pty-read","content":[{"type":"text","text":"dsh> printf 'PTY_OK\\n'\nPTY_OK\ndsh> \n[lines: 0-3 of 3]"}],"isError":false}],"role":"user","id":"116cdcf0-42dc-4220-8f01-38ed1153f903"}},"sourceEventSeqs":[35],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":3}} {"type":"step/start","data":{"turn":1,"step":4}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -43,9 +43,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"9bb8ec3d-6e6f-44a2-8957-8f2d855f4834"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[39,40,41,42,43],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b7216958-bc3a-4631-bf30-31db4f96b5f4"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[39,40,41,42,43],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":4,"callId":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}} -{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"pty-signal"},"content":[{"type":"tool-result","toolCallId":"pty-signal","content":[{"type":"text","text":"Error: unknown PTY session pty-missing"}],"isError":true}],"role":"user","id":"16fc1ca2-0eca-42d7-85d0-31794424c260"}},"sourceEventSeqs":[45],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"pty-signal"},"content":[{"type":"tool-result","toolCallId":"pty-signal","content":[{"type":"text","text":"Error: unknown PTY session pty-missing"}],"isError":true}],"role":"user","id":"d950a25e-05ca-4eec-aaf8-9937b8fe4f48"}},"sourceEventSeqs":[45],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":4}} {"type":"step/start","data":{"turn":1,"step":5}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -53,9 +53,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4ebdc957-0369-4bbb-a5a5-4d2ef8ac3493"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"6dae8392-a011-47e5-84d8-d9e4c3c19d0f"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":5,"callId":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}} -{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"pty-kill"},"content":[{"type":"tool-result","toolCallId":"pty-kill","content":[{"type":"text","text":"closed terminal session pty-1"}],"isError":false}],"role":"user","id":"55a3e2cc-dbc5-44bf-a824-e3bbc8568cd5"}},"sourceEventSeqs":[55],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"pty-kill"},"content":[{"type":"tool-result","toolCallId":"pty-kill","content":[{"type":"text","text":"closed terminal session pty-1"}],"isError":false}],"role":"user","id":"fb3d7081-eba7-4169-a2fe-e05ecae357ac"}},"sourceEventSeqs":[55],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":5}} {"type":"step/start","data":{"turn":1,"step":6}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -63,9 +63,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"6f956d23-5437-4a75-93a9-3abacd378e07"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[59,60,61,62,63],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b4305101-fbda-4a57-ac80-1c6cdae233f9"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[59,60,61,62,63],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":6,"callId":"pty-list","name":"terminal_list","arguments":"{}"}} -{"type":"tool/result","data":{"turn":1,"step":6,"message":{"source":{"kind":"tool","callId":"pty-list"},"content":[{"type":"tool-result","toolCallId":"pty-list","content":[{"type":"text","text":"(no terminal sessions)"}],"isError":false}],"role":"user","id":"be12d914-6fe1-4c1e-8262-65b2aa9c20e5"}},"sourceEventSeqs":[65],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":6,"message":{"source":{"kind":"tool","callId":"pty-list"},"content":[{"type":"tool-result","toolCallId":"pty-list","content":[{"type":"text","text":"(no terminal sessions)"}],"isError":false}],"role":"user","id":"0097f12a-a04e-4042-b47a-4a1535708049"}},"sourceEventSeqs":[65],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":6}} {"type":"step/start","data":{"turn":1,"step":7}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} @@ -73,6 +73,6 @@ {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":7,"message":{"role":"assistant","content":[{"type":"text","text":"DONE"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"9cca9680-5795-47d8-8edc-f6d44bcaa1ef"},"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[69,70,71,72,73],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":7,"message":{"role":"assistant","content":[{"type":"text","text":"DONE"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b75bf882-4d9d-4f92-9125-0cdc18f76f25"},"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[69,70,71,72,73],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":7}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/packages/workflow/tool-ralph/src/index.ts b/packages/workflow/tool-ralph/src/index.ts index 3e3e1b5943..d92d7251e2 100644 --- a/packages/workflow/tool-ralph/src/index.ts +++ b/packages/workflow/tool-ralph/src/index.ts @@ -406,7 +406,7 @@ export function apply(ctx: Context, config: Config): void { const resolved = resolveConfig(config) ctx.systemPrompt.section({ name: 'tool:ralph', - order: 116, + order: 117, text: 'Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.', }) ctx.tools.register(defineTool({ diff --git a/packages/workflow/tool-workflow/src/index.ts b/packages/workflow/tool-workflow/src/index.ts index 446c70ce03..899c6f8ab7 100644 --- a/packages/workflow/tool-workflow/src/index.ts +++ b/packages/workflow/tool-workflow/src/index.ts @@ -211,7 +211,7 @@ export function apply(ctx: Context, config: Config): void { // lives in tool plugins as prompt sections, not in the deployment persona). ctx.systemPrompt.section({ name: `tool:${toolName}`, - order: 115, + order: 116, text: `Use the ${toolName} tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.`, }) ctx.tools.register(defineTool({ From 25428f8e089a76cccf63ce4877e8827e55ba79b7 Mon Sep 17 00:00:00 2001 From: Yichen Jiang Date: Mon, 24 Aug 2026 14:12:36 +0800 Subject: [PATCH 2/2] fix(system-prompt): preserve downstream section order --- ...system-prompt-section-order-ties.i18n.yaml | 4 +-- ...-08-24-system-prompt-section-order-ties.md | 6 ++-- ...-24-system-prompt-section-order-ties.zh.md | 6 ++-- .../advanced-toolchain/session.1.jsonl | 2 +- .../advanced-toolchain/session.2.jsonl | 2 +- .../system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../both-mode-turn/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../code-mode-turn/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../lsp-definition/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../pty-tools/system-prompt.expected.md | 4 +-- .../read-image/system-prompt.expected.md | 4 +-- .../system-prompt.expected.md | 4 +-- .../system-prompt.1.expected.md | 4 +-- .../system-prompt.1.expected.md | 4 +-- .../session.1.jsonl | 2 +- .../session.2.jsonl | 2 +- .../system-prompt.1.expected.md | 4 +-- .../snapshots/subagent-mixed/session.1.jsonl | 2 +- .../snapshots/subagent-mixed/session.2.jsonl | 2 +- .../snapshots/subagent-multi/session.1.jsonl | 2 +- .../snapshots/subagent-multi/session.2.jsonl | 2 +- .../subagent-parallel/session.1.jsonl | 8 ++--- .../subagent-parallel/session.2.jsonl | 8 ++--- .../system-prompt.1.expected.md | 4 +-- .../text-turn/system-prompt.expected.md | 4 +-- .../web-fetch/system-prompt.expected.md | 4 +-- .../advanced-toolchain/session.1.jsonl | 10 +++--- .../advanced-toolchain/session.2.jsonl | 10 +++--- .../advanced-toolchain/session.jsonl | 28 +++++++-------- .../compaction-recovery/session.jsonl | 14 ++++---- .../tests/snapshots/pty-tools/session.jsonl | 34 +++++++++---------- packages/workflow/tool-ralph/src/index.ts | 2 +- packages/workflow/tool-workflow/src/index.ts | 2 +- 37 files changed, 108 insertions(+), 108 deletions(-) diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml index d53f6820ce..17d2684a6c 100644 --- a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.i18n.yaml @@ -2,5 +2,5 @@ # side as of the last confirmed-consistent state. Both languages carry equal authority; # after editing either side, bring the other along and re-record with: # pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md -2026-08-24-system-prompt-section-order-ties.md: 9aba5ee117a07b1d5abd6497232e61a48e55a639 -2026-08-24-system-prompt-section-order-ties.zh.md: 9736aa94f12a9ddf7242ef6d95b3c0204c4b62ed +2026-08-24-system-prompt-section-order-ties.md: 673c3b3cd668115ead9f9b5478c2bc432b78f930 +2026-08-24-system-prompt-section-order-ties.zh.md: 96a6843a0db48e52a2132ad9f8caf6243dfbbcd2 diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md index 9aba5ee117..673c3b3cd6 100644 --- a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.md @@ -10,7 +10,7 @@ English | [中文](2026-08-24-system-prompt-section-order-ties.zh.md) ## Decision -Give the three ordered sections distinct values: `tool:cordis` stays at 115, `tool:workflow` moves to 116, and `tool:ralph` moves from 116 to 117 so it remains after workflow. Prompt text and tool schemas remain unchanged. +Give the affected sequence distinct values without changing its established relative order: `tool:cordis` stays at 115, `tool:workflow` uses 115.5, `tool:ralph` stays at 116, continuable subagent guidance stays at 116.5, and child-report guidance stays at 117. Prompt text and tool schemas remain unchanged. ## Alternatives considered @@ -20,8 +20,8 @@ Give the three ordered sections distinct values: `tool:cordis` stays at 115, `to ## Consequences -The Cordis, workflow, and Ralph guidance has one platform-independent order. Prompt-section placements that require a stable relative position need distinct `order` values; stable sorting continues to preserve activation order for intentional ties. +The Cordis and workflow guidance has a platform-independent order while Ralph remains before continuable subagent and child-report guidance. Prompt-section placements that require a stable relative position need distinct `order` values; other equal-order sections retain activation-order semantics and are outside this decision. ## Testing -The keyless ACP and SDK snapshot replays pin the affected system-prompt order, and the full snapshot suite verifies the refreshed fixtures. +The keyless ACP and SDK snapshot replays pin Cordis before workflow and preserve the workflow, Ralph, continuable-subagent, and child-report sequence. The full snapshot suite verifies the refreshed fixtures. diff --git a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md index 9736aa94f1..96a6843a0d 100644 --- a/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md +++ b/.agents/notes/implemented/bug-fix/2026-08-24-system-prompt-section-order-ties.zh.md @@ -10,7 +10,7 @@ Status: implemented ## Decision -为这三个有序分段指定互不相同的 order：`tool:cordis` 保持 115，`tool:workflow` 移到 116，`tool:ralph` 从 116 移到 117，从而继续排在 workflow 之后。提示词文本与工具 schema 保持不变。 +在不改变既有相对顺序的前提下，为受影响的分段序列指定互不相同的 order：`tool:cordis` 保持 115，`tool:workflow` 使用 115.5，`tool:ralph` 保持 116，可继续运行的子代理指引保持 116.5，子代理报告指引保持 117。提示词文本与工具 schema 保持不变。 ## Alternatives considered @@ -20,8 +20,8 @@ Status: implemented ## Consequences -Cordis、workflow 与 Ralph 指引具有不依赖平台的唯一顺序。需要稳定相对位置的提示词分段必须使用互不相同的 `order`；对于有意设置的并列，稳定排序仍会保留激活顺序。 +Cordis 与 workflow 指引具有不依赖平台的顺序，同时 Ralph 仍排在可继续运行的子代理指引和子代理报告指引之前。需要稳定相对位置的提示词分段必须使用互不相同的 `order`；其他等序分段仍采用激活顺序，不属于本决策的范围。 ## Testing -无密钥 ACP 与 SDK 快照回放会固定受影响的系统提示词顺序，完整快照套件则验证刷新的 fixture。 +无密钥 ACP 与 SDK 快照回放会固定 Cordis 排在 workflow 之前，并保留 workflow、Ralph、可继续运行的子代理和子代理报告指引的顺序。完整快照套件验证刷新的 fixture。 diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl index 470217b161..f273ce4d5f 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Check direct child"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebe0cfa0-a909-47e0-8294-28ad84a8fe77"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"3d72b77d-f8f9-4e5b-b918-6ef0d13c35a2"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"fa9584c0-dfe6-4843-b9f0-41eaabf98ec0"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly DIRECT_CHILD_OK and","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl index a62cd266ab..052b62c82f 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"2ac2cc54-9bce-4cfa-a569-a64f51bc30a7"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"2c50b91f-f7ee-4aad-afc5-b08732633604"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"383e5a64-a889-40e5-a0bd-6504024c74b2"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly WORKFLOW_CHILD_OK and","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md index c20ff88a34..1743643d95 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md @@ -131,10 +131,10 @@ return { Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md index 4517bb46cb..7150bf2e6b 100644 --- a/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/agent-instructions/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md index 0c8262c90d..37df6287ed 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md @@ -25,10 +25,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md index 4913a69d8c..daf622df60 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-read-image/system-prompt.expected.md @@ -27,10 +27,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md index 79e1d537f8..6894f13fb6 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md @@ -27,10 +27,10 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + ## Writing code for run_code `run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program: diff --git a/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md index 95936c6f15..9b4698844c 100644 --- a/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/fs-glob-sampling/system-prompt.expected.md @@ -18,6 +18,6 @@ Use the web_search tool to discover current information on the web. The required Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md index fee88385ae..b906b6f3c8 100644 --- a/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/lsp-definition/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md index f37a2d570a..545e903230 100644 --- a/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/product-subagent-codex/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md index 9939935412..06b614520c 100644 --- a/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/pty-tools/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md index dd0ebca77d..a0d3386eaa 100644 --- a/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/read-image/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md index 5166d0236f..800356dccc 100644 --- a/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/session-query-spill/system-prompt.expected.md @@ -27,6 +27,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md index 2da38d57c5..b198b48a12 100644 --- a/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-continuable-inheritance/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md index 2da38d57c5..b198b48a12 100644 --- a/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-continuable/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl index 2e7f6f7c9a..7c6db15da2 100644 --- a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Start depth one"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Call subagent once. Ask that child to attempt one further subagent call, then report the result."}],"source":{"kind":"user"},"role":"user","id":"a8129357-1bde-4cbd-90b4-6b8ad51d52e1"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"37b2a8d1-90f2-4aad-9e35-737559cdaaa2"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"cb8e46e2-6f46-4f4c-ab97-96b925df62b4"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Call subagent once. Ask that","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl index 3a3d440f2d..1b65465994 100644 --- a/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-depth-two-rejection/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Start depth two"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Attempt one subagent call beyond the configured cap, then report the rejection."}],"source":{"kind":"user"},"role":"user","id":"d4dc5a16-e542-4dd9-8e82-e6b7829cfc4b"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"776d4567-6a92-45a6-8a74-8d376af180e5"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"45e7b5cb-cd03-4a92-9560-09317f5c3e15"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Attempt one subagent call beyond","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md index 2da38d57c5..b198b48a12 100644 --- a/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-list-agents/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl index 474ccadf60..df79bc0066 100644 --- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Reply ALPHA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"73ce401a-faaf-408a-879e-7485380d537d"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"c3bc9e3e-bf25-4169-ad2c-bda0183fad84"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"41fe3733-7671-416c-affa-a71500ef0936"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl index 8ff90306d3..d55d6f34e8 100644 --- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl @@ -31,7 +31,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"fork","label":"Recall project codeword"}} {"type":"step/start","data":{"turn":2,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else."}],"source":{"kind":"user"},"role":"user","id":"86e9f144-764f-460d-b72b-262cffe43d77"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"95dee7dd-6ac8-4695-b3f0-745cb724dfbd"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"57e307d9-3800-41f4-80b3-511476d84745"},"surfaceOp":"append"} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}} {"type":"assistant/chunk","data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} {"type":"reasoning-chunks","data":{"turn":2,"step":1,"index":0,"dt":[0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"texts":["The"," user"," is"," asking"," me"," to"," recall"," the"," project"," cod","ew","ord"," that"," was"," mentioned"," earlier"," in"," the"," conversation","."," I"," was"," told"," to"," remember"," it",":"," SA","FF","RON","."]}} diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl index 6cea54fb30..750bc35d2a 100644 --- a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Return ALPHA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"a287f842-f6f2-4a17-ab4c-820e41f498d5"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"967aa8c4-7ad8-4973-966a-a6fcf8ba00dd"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"0018c379-5d56-4594-b980-1a6009ef8767"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl index 2888c8f591..05ded9867f 100644 --- a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl @@ -8,7 +8,7 @@ {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Return BETA only"}} {"type":"step/start","data":{"turn":1,"step":1}} {"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word BETA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"53f6419d-8ddc-4eee-8803-5b68411336f9"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"c52ae99e-3f24-4803-a002-56564ed59845"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"3d741097-e063-4fc1-bba0-1f7f1dc4fc45"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} diff --git a/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl index 0cb0df4130..734566070e 100644 --- a/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-parallel/session.1.jsonl @@ -2,19 +2,19 @@ {"type":"sandbox/mode","data":{"mode":"danger-full-access","source":"delegation"}} {"type":"approval/policy","data":{"policy":"never","source":"delegation"}} {"type":"permission/preset","data":{"preset":"danger-full-access"}} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c0a22dc3-7ed8-401f-9d27-daa216035e79"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"9e5dc791-bb17-4354-bbdc-41b8a4ad6b17"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Say the word ALPHA"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"c0a22dc3-7ed8-401f-9d27-daa216035e79"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"f995fac5-e34b-468d-9162-d64cbac11ab8"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"9e5dc791-bb17-4354-bbdc-41b8a4ad6b17"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"2e5f49b2-e459-47dd-b347-9c5b68c244de"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ALPHA"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"2814fde1-ee68-4d3c-a14f-7b9ea936384c"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"df5873d9-8eba-49b9-9634-0231b3945982"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl index 86ae5204d3..23b85f753c 100644 --- a/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl +++ b/examples/acp-agent/tests/snapshots/subagent-parallel/session.2.jsonl @@ -2,19 +2,19 @@ {"type":"sandbox/mode","data":{"mode":"danger-full-access","source":"delegation"}} {"type":"approval/policy","data":{"policy":"never","source":"delegation"}} {"type":"permission/preset","data":{"preset":"danger-full-access"}} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"6bbb164f-99ee-4b7f-a0bd-c1a35b02a076"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"0fb40673-0440-4ed3-ba2f-07eac9586e05"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Say the word ALPHA"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"6bbb164f-99ee-4b7f-a0bd-c1a35b02a076"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"7cbd146e-97ac-49f2-979a-4221d9277ec8"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"},"role":"user","id":"0fb40673-0440-4ed3-ba2f-07eac9586e05"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations.\n\nApproval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."},{"name":"approval:policy","text":"Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`)."},{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"8394d04c-a143-4b26-867f-3e4e04a94116"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly the word","messageSeqs":[8],"source":{"kind":"fallback"}}} {"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ALPHA"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"c40ec82f-7cb8-4885-be04-ba85bff6e2a2"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"ALPHA"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"c31a0818-10e0-471d-afe1-8ff2fee62eda"}},"sourceEventSeqs":[13,14,15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md b/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md index 2da38d57c5..b198b48a12 100644 --- a/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md +++ b/examples/acp-agent/tests/snapshots/subagent-report/system-prompt.1.expected.md @@ -25,8 +25,8 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. + Deliver your result with the report tool before you finish: call it once with a self-contained answer. The agent that started you shares your workspace but does not automatically receive your transcript, tool output, or reasoning, so a closing remark such as "done" leaves it nothing it can use. Report earlier as well whenever a partial finding changes what that agent should do next; reporting never ends your turn. diff --git a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md index b1d7332947..975b5a7baf 100644 --- a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md index ab93903173..b70cc036d4 100644 --- a/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/web-fetch/system-prompt.expected.md @@ -25,6 +25,6 @@ Use goal tools for one long-running completion objective in the current session. Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls. -Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. - Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out. + +Use subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message. diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl index b1e3b0e457..100f729a7e 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.1.jsonl @@ -1,19 +1,19 @@ {"type":"session","version":0,"id":"22222222-2222-4222-8222-222222222222","createdAt":1783950001000,"cwd":"{{cwd}}","parentSession":"11111111-1111-4111-8111-111111111111","origin":"subagent","delegationDepth":1} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"f9b76611-65ac-4d61-ba91-cb003d56670b"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"94e03719-effb-4b99-bba9-621d1798a933"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn","label":"Check direct child"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"f9b76611-65ac-4d61-ba91-cb003d56670b"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"09b80e2e-0a74-48b9-88c3-03e1d859fb5d"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"94e03719-effb-4b99-bba9-621d1798a933"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"3f567eec-b0d3-4692-9e89-bd2a53623b69"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly DIRECT_CHILD_OK and","messageSeqs":[5],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"DIRECT_CHILD_OK"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DIRECT_CHILD_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"f018f8fa-7a81-46a3-93a0-a30705a7785f"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4c10875e-5b61-4fbf-81e4-6306ab793f77"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl index 6be8abb48f..dd372221a9 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.2.jsonl @@ -1,19 +1,19 @@ {"type":"session","version":0,"id":"33333333-3333-4333-8333-333333333333","createdAt":1783950002000,"cwd":"{{cwd}}","parentSession":"11111111-1111-4111-8111-111111111111","origin":"subagent","delegationDepth":1} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebb3764d-0742-4054-a421-69d479b90021"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"335b85c1-27e6-4047-b7d6-24ccc388ea13"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"subagent/descriptor","data":{"version":2,"mode":"one-shot","provider":"spawn"}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"ebb3764d-0742-4054-a421-69d479b90021"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"8390dc9d-26c3-439b-9f60-ac263964f0d3"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"},"role":"user","id":"335b85c1-27e6-4047-b7d6-24ccc388ea13"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nYou are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"subagent:delegation","text":"You are a delegated subagent: your permission scope was fixed when you were started and cannot be widened from inside this session — operations that require approval are rejected automatically. When the task needs access beyond that scope, do not retry the denied operation; state the limitation in your reply so the delegating agent can handle it."}]},"role":"user","id":"1ae21e84-a121-420a-a80a-986bcb9bfd0a"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Reply with exactly WORKFLOW_CHILD_OK and","messageSeqs":[5],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"WORKFLOW_CHILD_OK"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"WORKFLOW_CHILD_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"80f282de-e531-41cc-94d9-d6896f6a2800"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b72bbf47-34ca-4fa1-bb18-37d0fb6150be"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[10,11,12,13,14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl index dbe2d47352..933d98387f 100644 --- a/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl +++ b/examples/headless-agent/tests/snapshots/advanced-toolchain/session.jsonl @@ -1,20 +1,20 @@ {"type":"session","version":0,"id":"11111111-1111-4111-8111-111111111111","createdAt":1783950000000,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"be704496-d340-4e19-8506-6cf81feb739a"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"587f027c-803f-4141-b6f9-820b6b1379ea"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"be704496-d340-4e19-8506-6cf81feb739a"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Run this advanced flow exactly once: define a host-only dynamic Cordis Package named Snapshot Marker; run and inspect snap-1/pkg-1 through run_code; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; remove snap-1; then reply with exactly ADVANCED_HEADLESS_OK."}],"source":{"kind":"user"},"role":"user","id":"587f027c-803f-4141-b6f9-820b6b1379ea"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Run this advanced flow exactly","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.\n\nVerify your work by running the code or tests. Keep answers brief and factual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\n# Dynamic Cordis Plugins\n\nDynamic Cordis plugins temporarily extend the current DSH process. A Plugin uses apply(ctx) to consume Services, listen to Events, provide Services, register model Tools, or register browser UI in Slots.\n\n- Plugin and Package definitions exist only in the current process. define itself does not modify repository source, configuration, or disk, and definitions do not survive a process restart.\n- The restricted execution environment prevents accidental misuse; it is not a security boundary for malicious code. Services obtained by dynamic code connect to the real runtime.\n\n## Make the user-facing plan clear first\n\n- Dynamic Cordis Plugins are one available implementation mechanism, not the default for every request. Consider whether one could help only when the user intends to design or create something, or when a temporary interface could materially aid the current work. The presence of these instructions or Tools, and discussion of Cordis itself, do not make a request a dynamic-Plugin task.\n- When Cordis is a plausible fit, infer the intended work target and lifetime from the request and conversation. Use it only when the outcome belongs to the current running harness and should be delivered as a temporary runtime extension. If that distinction is materially ambiguous, ask at most one concise question about the intended result or lifetime. Otherwise proceed with the matching workflow; do not require the user to know or choose Cordis as an implementation mechanism.\n- Once a dynamic Plugin is appropriate, decide whether the task creates a new Plugin or modifies the Plugin named by the user with @pluginId. Proceed directly when the goal is clear; do not ask for repeated confirmation.\n- Choose Host, Client, or both from the requested outcome. Do not propose a Client/browser UI when the task does not need visible page behavior, and do not avoid Client when the requested outcome is visual, interactive, or depends on page state. Host versus Client is an implementation choice; do not make the user choose it.\n- When a design direction or a potentially useful interface would materially affect the result, ask at most one concise outcome or creative-preference question and offer a few candidate directions. Otherwise proceed directly; do not conduct a multi-round interview or a complex questionnaire.\n- cordis_define only defines and presents code; it does not run it. After definition, explain the pluginId and packageId returned by the Host and whether the next step is a run or update.\n- cordis_run may require user approval. When it returns awaiting-approval, explain that the user must allow or reject it in the UI. Do not wait, retry, or claim that it is running.\n- When it returns starting, explain that the request has entered the asynchronous flow and the Client is still activating. starting does not mean success. Wait for the system to report the final result through steering context.\n- Do not request approval again after the user rejects it. After a technical failure, fix the same Plugin from its diagnostics; do not silently create a replacement Plugin.\n\n## Recommended workflow and Tools\n\nBefore creating, modifying, or repairing a Plugin, load the cordis-plugin-development Skill. The Skill provides requirement navigation, capability composition, complete examples, and troubleshooting. Treat Inspect Provider results as the source of truth for exact APIs.\n\n1. cordis_inspect_list: discover the current Host and Client Providers and their read-only query methods.\n2. cordis_inspect_query: use the returned platform, provider, method, and schema to query exact Service, Event, Builtin, Slot, Theme token, or Tool information.\n3. cordis_inspect_self: inspect the current Session's Plugins, Packages, version pointers, source, and diagnostics. Source is returned only when both pluginId and packageId are specified.\n4. cordis_define: create the first Package for a new Plugin or append an immutable Package to an existing Plugin. It defines code but does not run it.\n5. cordis_run: activate an exact Package. Use run for the first activation, restarting current, or rollback; use update to switch versions.\n6. cordis_stop: remove the current Run and pending approval request while retaining definitions, grants, and version pointers.\n7. cordis_undefine: permanently stop and delete a Plugin and all of its Packages. Use it only after confirming that the user no longer needs them.\n\n- Inspect and Catalog data only confirm capabilities, names, signatures, types, and registration protocols before code is written; they do not replace business APIs.\n- Query Service.listService and Event.listEvents without input to choose from their compact signature directories, then query the exact service or event before using it. Exact queries return the structured contract and only its referenced types.\n- At runtime, a Plugin must call real Services or listen to real Events. Do not cache, display, or depend on Inspect results as business data.\n\n## Identity, versions, and approval\n\n- pluginId identifies a Plugin that can be modified over time. For a new Plugin, submit only a semantic idPrefix of 3–6 lowercase English letters; the Host allocates the final ID.\n- packageId identifies one immutable Host/Client source version under a Plugin. To change code, define a new Package; never overwrite an old version.\n- pluginRunId identifies one activation attempt and connects its approval, Host/Client loading, private RPC, Run card, and errors.\n- currentPackageId is the most recent fully successful Package. Stopping, starting an update, or failing an update does not clear it.\n- nextPackageId is the target awaiting approval, being attempted, awaiting Client activation, or most recently failed.\n- A single check mark authorizes only the current Package; double check marks authorize future versions of the same Plugin. A grant remains in effect after a technical failure.\n- An update stops the old Run before starting the target Package. Failure does not automatically restart the old version; retry next with update or roll back to current with run.\n\nWhen the user enters @pluginId, the system injects identity, the default base Package, version pointers, and runtime status, but not source code:\n\n1. Call cordis_inspect_self(pluginId, packageId) to read the target source.\n2. Use cordis_define in existing mode to append a Package to the same Plugin.\n3. Call cordis_run in run or update mode according to the version relationship.\n\nNever silently create another Plugin for @pluginId. If the reference is unavailable because it was removed, belongs to another Session, or was lost on process restart, tell the user directly.\n\n## High-frequency errors that must be avoided\n\n### Services: ctx.get and inject\n\n- Read an optional Service with ctx.get('serviceName') by default and handle undefined.\n- Declare inject: ['serviceName'] on the returned Plugin object only when the Service is a hard dependency and the Plugin must enter waiting until Cordis reactivates it after the Service appears.\n- Read ctx.serviceName only after declaring that Service in inject. Never access an undeclared Service as a ctx property.\n\n```js\nreturn {\n inject: ['requiredService'],\n apply(ctx) {\n ctx.requiredService.someMethod()\n const optionalService = ctx.get('optionalService')\n if (optionalService !== undefined) optionalService.someMethod()\n },\n}\n```\n\n### Code: use plain JavaScript only\n\n- Host and Client code is not transformed by TypeScript, JSX, or a bundler.\n- Do not use TypeScript types, as, decorators, import, require, or JSX.\n- Client React code must use React.createElement(...); never write .\n- Do not assume that process, Buffer, window, document, fetch, native timers, or any other global is available. Query the corresponding platform's Builtins and Services first.\n\n### Data: do not serialize live data\n\n- Services, Events, Slots, Sessions, and their derived Cordis/DSH objects are internal live data, not ordinary JSON that can be dumped.\n- Do not apply JSON.stringify, structuredClone, recursive enumeration, full copying, or whole-object display to live data.\n- Read only the leaf fields required by the task, then construct the smallest owned data object without Host references.\n\n### Lifecycle: every side effect must be reversible\n\n- Services, Events, Tools, handlers, timers, Slots, styles, and theme overrides must all belong to the current Fiber.\n- Use ctx.effect(), ctx.on(), or official APIs that return a disposer so stop, update, or undefine removes every side effect.\n- The cordis-plugin-development Skill contains complete timer, Waterfall, Slot, theme, Tool, RPC, and React examples and troubleshooting guidance.\n\n## Host and Client\n\n- Host runs in the DSH Node.js process and is appropriate for files, networking, commands, Agent/Session access, Host Events, Services, model Tools, and JSON methods callable by the Client.\n- Client runs in the browser page and is appropriate for themes, layout, current page state, Tool cards, and Slot UI.\n- Host and Client communicate through Package-private JSON methods: Host uses harness.handle(method, handler), and Client uses host.call(method, args). The direction is Client→Host, and only lossless JSON may cross it.\n- Client UI must be registered in a queried Slot; apply() cannot directly return a React Element. Query Slots.listSubTree without root to choose from the compact purpose/topology tree, then query the exact root for its full registration contract and props before writing code.\n- See the Skill and Inspect Providers for Run-specific panels and exact Slot registration patterns.\n\n## Asynchronous results and recovery\n\n- Do not wait inside a Tool for approval or browser work that can happen only after the current turn ends.\n- Asynchronous success, rejection, and runtime errors update Run state and notify you through steering context.\n- After a technical failure, use cordis_inspect_self to read the exact Package source and its message/stack. Define a corrected Package under the same Plugin and retry autonomously.\n- Use the cordis-plugin-development Skill for other failure causes, repair procedures, and complete extension patterns.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\n## Writing code for run_code\n\n`run_code` takes two required arguments: `code` — the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped) — and `description`, a short summary of what the program does. Inside the program:\n\n- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools[\"my-tool\"](args)`. Every call resolves to the tool's typed canonical JSON value. Tool arguments must be lossless JSON.\n- A FAILED tool call rejects with `ToolCallError`, whose `toolName` identifies the failed tool and whose `message` is human-readable — `try/catch` it to handle and continue.\n- Independent read-only calls MAY overlap under `Promise.all` (safe calls run concurrently; mutating calls run alone, in submission order). Sequence dependent work with `await`.\n- Emit results with `return` and/or `console.log(...)`. Only what you print or return is program output. A successful tool result containing an image is attached after the run so you can inspect it on the next step; every other intermediate result stays out of the conversation, so extract just what you need.\n\nThe available tools:\n\n```ts\ntype JsonValue = null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue }\n\ninterface ToolArgsMap {\n /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`. */\n bash: {\n /** The bash command to execute. */\n command: string;\n /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\". */\n description: string;\n /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */\n timeoutMs?: number;\n /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */\n workdir?: string;\n /** Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies. */\n run_in_background?: boolean;\n } & Record;\n /** Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs. */\n cordis_define: {\n plugin: {\n kind: \"new\";\n /** Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix. */\n idPrefix: string;\n } | {\n kind: \"existing\";\n /** Exact ID of an existing Plugin; the new Package is appended to that instance. */\n pluginId: string;\n };\n /** Short, readable Package name. */\n name: string;\n /** One-sentence, user-facing description of the Package purpose. */\n purpose: string;\n code: {\n /** Plain JavaScript function body that returns the Host-half Cordis Plugin. */\n host?: string;\n /** Plain JavaScript function body that returns the browser Client-half Cordis Plugin. */\n client?: string;\n };\n } & Record;\n /** List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call. */\n cordis_inspect_list: Record;\n /** Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props. */\n cordis_inspect_query: {\n /** Runtime platform that owns the Provider. */\n platform: \"host\" | \"client\";\n /** Exact Provider ID returned by cordis_inspect_list. */\n provider: string;\n /** Exact method name declared by the Provider manifest. */\n method: string;\n /** Optional query input; it must satisfy the method input schema. */\n input?: JsonValue;\n } & Record;\n /** Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers. */\n cordis_inspect_self: {\n /** Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin. */\n pluginId?: string;\n /** Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned. */\n packageId?: string;\n } & Record;\n /** Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it. */\n cordis_run: {\n /** Stable Plugin ID returned by cordis_define. */\n pluginId: string;\n /** Exact immutable Package ID to activate under that Plugin. */\n packageId: string;\n /** Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package. */\n mode: \"run\" | \"update\";\n } & Record;\n /** Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal. */\n cordis_stop: {\n /** Stable dynamic Plugin ID to stop. */\n pluginId: string;\n } & Record;\n /** Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead. */\n cordis_undefine: {\n /** Stable dynamic Plugin ID to remove permanently. */\n pluginId: string;\n } & Record;\n /** Edit an existing UTF-8 text file by replacing literal text. */\n edit: {\n /** Path to edit, resolved by the filesystem backend. */\n file_path: string;\n /** Literal text to replace. Must match exactly. */\n old_string: string;\n /** Literal replacement text. Use an empty string to delete the match. */\n new_string: string;\n /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */\n replace_all?: boolean;\n } & Record;\n /** Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op. */\n interrupt_agent: {\n /** The agent id of the running agent to interrupt. */\n agent_id: string;\n } & Record;\n /** Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops. */\n job_kill: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Optional short reason, recorded in the log and forwarded to the job. */\n reason?: string;\n } & Record;\n /** List your background jobs (running and finished) with their ids, kinds, and statuses. */\n job_list: Record;\n /** Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */\n job_output: {\n /** Job id returned by the tool that started the background work. */\n job_id: string;\n /** Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive. */\n wait?: boolean;\n /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */\n timeout_ms?: number;\n } & Record;\n /** Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools. */\n ralph: {\n /** The immutable completion objective for every fresh Ralph round. */\n objective: string;\n /** Optional positive safe-integer round cap, bounded by the deployment ceiling. */\n maxRounds?: number;\n } & Record;\n /** Read a UTF-8 text file and return line-numbered content. */\n read: {\n /** Path to read, resolved by the filesystem backend. */\n file_path: string;\n /** 1-based first line to return. Defaults to 1. */\n offset?: number;\n /** Maximum number of lines to return. Defaults to 2000. */\n limit?: number;\n } & Record;\n /** Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered. */\n send_message: {\n /** The subagent id returned when the background subagent was started. */\n subagent_id: string;\n /** The message to deliver to the subagent. */\n message: string;\n } & Record;\n /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */\n skill: {\n /** The exact skill name from the available skills list. */\n name: string;\n } & Record;\n /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result. */\n subagent: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */\n prompt: string;\n /** Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it. */\n run_in_background?: boolean;\n } & Record;\n /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result. */\n subagent_fork: {\n /** A short (3-5 word) description of the delegated task, for display. */\n description: string;\n /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */\n prompt: string;\n } & Record;\n /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */\n todo_write: {\n /** The COMPLETE task list, replacing any previous list. */\n todos: ({\n /** What the task is — a short imperative line. */\n content: string;\n /** pending (not started) | in_progress (now) | completed (done). */\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n } & Record;\n /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */\n workflow: {\n /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `). */\n script: string;\n /** The workflow identity block (plain JSON — never code). */\n meta: {\n /** Short kebab-case workflow name. */\n name: string;\n /** One-line description of what the workflow does. */\n description: string;\n /** Optional guidance on when this workflow applies. */\n whenToUse?: string;\n /** Optional phase declarations matched by phase() calls. */\n phases?: ({\n /** The phase title phase() calls match by exact string. */\n title: string;\n /** Optional one-line description of the phase. */\n detail?: string;\n /** Optional provider override this phase is expected to use. */\n provider?: string;\n /** Optional model override this phase is expected to use. */\n model?: string;\n } & Record)[];\n } & Record;\n /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}). */\n args?: Record;\n } & Record;\n /** Create or fully replace a UTF-8 text file. */\n write: {\n /** Path to write, resolved by the filesystem backend. */\n file_path: string;\n /** Full UTF-8 text content to write. */\n content: string;\n } & Record;\n}\n\ninterface ToolOutputMap {\n bash: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"foreground\";\n exitCode: number | null;\n signal: string | null;\n timedOut: boolean;\n aborted: boolean;\n timeoutMs: number;\n stdout: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n stderr: {\n text: string;\n truncated: boolean;\n spillPath?: string;\n };\n sandbox?: {\n mode: string;\n denied: boolean;\n enforcement?: string;\n runnerFailed?: boolean;\n };\n };\n cordis_define: {\n pluginId: string;\n packageId: string;\n name: string;\n purpose: string;\n hasHostHalf: boolean;\n hasClientHalf: boolean;\n };\n cordis_inspect_list: JsonValue;\n cordis_inspect_query: JsonValue;\n cordis_inspect_self: JsonValue;\n cordis_run: JsonValue;\n cordis_stop: {\n pluginId: string;\n };\n cordis_undefine: {\n pluginId: string;\n wasRunning: boolean;\n };\n edit: {\n path: string;\n before: string;\n after: string;\n };\n interrupt_agent: {\n accepted: boolean;\n };\n job_kill: {\n outcome: \"cancellation-requested\" | \"already-finished\";\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n job_list: ({\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n })[];\n job_output: {\n text: string;\n job: {\n id: string;\n kind: string;\n label: string;\n status: \"running\" | \"stopping\" | \"completed\" | \"killed\" | \"failed\";\n detail?: string;\n startedAt: number;\n finishedAt?: number;\n };\n };\n ralph: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n read: {\n path: string;\n offset: number;\n lines: {\n number: number;\n text: string;\n }[];\n totalLines: number;\n };\n send_message: {\n messageId: string;\n };\n skill: {\n name: string;\n provider: string;\n resourceBase?: {\n kind: \"directory\";\n path: string;\n } | {\n kind: \"url\";\n url: string;\n } | {\n kind: \"opaque\";\n description: string;\n };\n content: string;\n };\n subagent: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n subagent_fork: {\n kind: \"background\";\n jobId: string;\n } | {\n kind: \"continuable\";\n subagentId: string;\n } | {\n kind: \"foreground\";\n runId: string;\n output: JsonValue[];\n };\n todo_write: {\n todos: ({\n content: string;\n status: \"pending\" | \"in_progress\" | \"completed\";\n })[];\n counts: {\n pending: number;\n inProgress: number;\n completed: number;\n };\n };\n workflow: {\n runId: string;\n agentsStarted: number;\n result: JsonValue;\n };\n write: {\n path: string;\n operation: \"create\" | \"update\";\n before: string | null;\n after: string;\n };\n}\n\ntype ToolName = keyof ToolOutputMap\n\ndeclare class ToolCallError extends Error {\n readonly name: \"ToolCallError\";\n readonly toolName: ToolName;\n}\n\ndeclare const tools: {\n [K in ToolName]: (args: ToolArgsMap[K]) => Promise;\n}\n```","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"cordis_define","description":"Define an immutable Cordis Package. For a new Plugin, use kind:\"new\" and provide only a semantic prefix of 3–6 lowercase English letters; the Host returns the final pluginId and packageId. To modify an existing Plugin, use kind:\"existing\" with its exact pluginId to append a Package without overwriting older versions. Provide at least one of code.host and code.client. Each value is a plain JavaScript function body that returns a Cordis Plugin; no TypeScript, JSX, or import transformation occurs. Query Inspect before depending on a Service, Event, Builtin, Slot, or token. Define only validates parameters and syntax and records source: it does not request approval, execute apply, or change currentPackageId. On success, call cordis_run with the returned IDs.","parameters":{"type":"object","properties":{"plugin":{"oneOf":[{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"new"},"idPrefix":{"type":"string","description":"Suggested semantic prefix of 3–6 lowercase English letters; the Host adds a unique numeric suffix."}},"required":["kind","idPrefix"]},{"type":"object","additionalProperties":false,"properties":{"kind":{"type":"string","const":"existing"},"pluginId":{"type":"string","description":"Exact ID of an existing Plugin; the new Package is appended to that instance."}},"required":["kind","pluginId"]}]},"name":{"type":"string","description":"Short, readable Package name."},"purpose":{"type":"string","description":"One-sentence, user-facing description of the Package purpose."},"code":{"type":"object","additionalProperties":false,"properties":{"host":{"type":"string","description":"Plain JavaScript function body that returns the Host-half Cordis Plugin."},"client":{"type":"string","description":"Plain JavaScript function body that returns the browser Client-half Cordis Plugin."}}}},"required":["plugin","name","purpose","code"]}},{"name":"cordis_inspect_list","description":"List every Cordis Inspect Provider currently known to the Host, including local Host Providers and the latest manifests synchronized from the Client. Each entry includes its platform, purpose, read-only methods, and input/output schemas. Call this Tool before creating or modifying a Package, then select the provider and method for cordis_inspect_query from its result. Do not guess names or treat an Inspect method as a business Service that Plugin code can call.","parameters":{"type":"object","properties":{}}},{"name":"cordis_inspect_query","description":"Run a read-only query explicitly declared by an Inspect Provider. platform, provider, and method must come from cordis_inspect_list, and input must satisfy that method's schema. Use this Tool before cordis_define to read exact Service methods, Event modes, Builtin signatures, Tool schemas, theme tokens, or live Slot trees and props. Host queries run locally. A Client query waits for the first valid page response and remains pending until a page answers or the Tool is cancelled. This Tool cannot invoke business Service methods or modify the runtime. For Service.listService and Event.listEvents, query without input to navigate the compact signature directory, then query the exact service or event for its structured contract and referenced types. For Slots.listSubTree, query without root to navigate the compact tree, then query the exact root for its complete registration contract and props.","parameters":{"type":"object","properties":{"platform":{"type":"string","description":"Runtime platform that owns the Provider.","enum":["host","client"]},"provider":{"type":"string","description":"Exact Provider ID returned by cordis_inspect_list."},"method":{"type":"string","description":"Exact method name declared by the Provider manifest."},"input":{"description":"Optional query input; it must satisfy the method input schema."}},"required":["platform","provider","method"]}},{"name":"cordis_inspect_self","description":"Inspect dynamic Cordis objects owned by the current Session at increasing levels of detail. With no IDs, list only Plugin summaries. With pluginId alone, return version pointers, the latest Run, and every Package summary. Only pluginId plus packageId returns that immutable Package's Host/Client source and runtime diagnostics. packageId cannot be supplied alone. Query an exact Package before handling @pluginId, repairing an asynchronous failure, or defining an updated version. This Tool is read-only: it neither executes code nor changes version pointers.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define or injected by @pluginId; omit it to list every current Plugin."},"packageId":{"type":"string","description":"Exact immutable Package ID owned by pluginId; when specified, source and diagnostics are returned."}}}},{"name":"cordis_run","description":"Activate one exact Package of a dynamic Plugin. Use mode:\"run\" for the first activation, restarting currentPackageId, or rollback. When current exists, use mode:\"update\" to switch to a different Package, even if the Plugin is currently stopped. An unauthorized Client Package creates an approval request and returns awaiting-approval; an authorized Package returns starting and continues asynchronously in the browser. Neither result waits for the final outcome inside the Tool. currentPackageId changes only after complete success; on failure, the old current and target next remain. Asynchronous success, rejection, or technical failure is reported through state and steering. After a technical failure, read diagnostics with cordis_inspect_self, correct the same Plugin, and retry autonomously. Do not request approval again after the user rejects it.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable Plugin ID returned by cordis_define."},"packageId":{"type":"string","description":"Exact immutable Package ID to activate under that Plugin."},"mode":{"type":"string","description":"Use run for the first activation, restarting current, or rollback; use update to switch from current to a different Package.","enum":["run","update"]}},"required":["pluginId","packageId","mode"]}},{"name":"cordis_stop","description":"Stop the current Run of a dynamic Plugin and cancel unfinished approval or activation requests. Retain the Plugin, every immutable Package, grants, currentPackageId, and nextPackageId so it can later run or update directly. Stopping an already stopped Plugin succeeds idempotently. Use this Tool to disable effects temporarily; use cordis_undefine for permanent removal.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to stop."}},"required":["pluginId"]}},{"name":"cordis_undefine","description":"Permanently remove a dynamic Plugin owned by the current Session. If it is running or awaiting approval, first stop it and cancel the request, then delete every Package, grant, and version pointer. After this returns, its pluginId, packageIds, @ reference, and Package business views are invalid; historical cards retain only a \"Plugin removed\" record. Do not call this Tool when versions must remain available for restart or rollback; use cordis_stop instead.","parameters":{"type":"object","properties":{"pluginId":{"type":"string","description":"Stable dynamic Plugin ID to remove permanently."}},"required":["pluginId"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"run_code","description":"Execute a TypeScript program against the available tools. Takes two required arguments: `code`, the BODY of an async function (erasable syntax only; top-level `await` and `return` work), and `description`, a short summary of what the program does. Call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return is program output — curate it. Image-bearing subtool results are attached after the run.","parameters":{"type":"object","properties":{"code":{"type":"string","description":"The program: the body of an async TypeScript function."},"description":{"type":"string","description":"Clear, concise description of what this program does in active voice, 5-10 words (shown in the UI). Examples: \"Count TODO markers across packages\"; \"Read failing test and its fixture\"; \"Rename config key in every cordis.yml\"."}},"required":["code","description"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"advanced-define","name":"cordis_define","argumentsDelta":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"28ed15b0-0b10-4cb7-bdb8-17af0c101f9a"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"9360e8c1-8782-489a-abe7-7b2900b3e43e"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"advanced-define","name":"cordis_define","arguments":"{\"plugin\":{\"kind\":\"new\",\"idPrefix\":\"snap\"},\"name\":\"Snapshot Marker\",\"purpose\":\"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\"code\":{\"host\":\"return { apply() {} }\"}}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"advanced-define"},"content":[{"type":"tool-result","toolCallId":"advanced-define","content":[{"type":"text","text":"Defined snap-1/pkg-1 (Snapshot Marker); it is not running yet. Use cordis_run to activate this Package."}],"isError":false}],"role":"user","id":"f037d5e8-3ef3-4c76-8bd9-e823543d16c2"},"meta":{"pluginId":"snap-1","packageId":"pkg-1"}},"sourceEventSeqs":[14],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"advanced-define"},"content":[{"type":"tool-result","toolCallId":"advanced-define","content":[{"type":"text","text":"Defined snap-1/pkg-1 (Snapshot Marker); it is not running yet. Use cordis_run to activate this Package."}],"isError":false}],"role":"user","id":"3cccd752-2d82-4e76-8aeb-65f8181f5d67"},"meta":{"pluginId":"snap-1","packageId":"pkg-1"}},"sourceEventSeqs":[14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -22,13 +22,13 @@ {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"8a02e55c-d6e7-4b63-9674-856c71c385a1"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[18,19,20,21,22],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"f57879b3-f552-4209-9d7b-9b8eb14311b3"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[18,19,20,21,22],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":2,"callId":"advanced-code","name":"run_code","arguments":"{\"code\":\"const run = await tools.cordis_run({ pluginId: 'snap-1', packageId: 'pkg-1', mode: 'run' });\\nconst inspected = await tools.cordis_inspect_self({ pluginId: 'snap-1' });\\nreturn { run, inspected };\",\"description\":\"Run and inspect the dynamic Cordis Package\"}"}} {"type":"tool/code-dispatch-start","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_run","arguments":{"pluginId":"snap-1","packageId":"pkg-1","mode":"run"}}} {"type":"tool/code-dispatch","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_run","arguments":{"pluginId":"snap-1","packageId":"pkg-1","mode":"run"},"isError":false,"content":[{"type":"text","text":"snap-1/pkg-1 is running (run-1)."}]}} {"type":"tool/code-dispatch-start","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:2","name":"cordis_inspect_self","arguments":{"pluginId":"snap-1"}}} {"type":"tool/code-dispatch","data":{"rootCallId":"advanced-code","parentCallId":"advanced-code","subCallId":"advanced-code:code:2","name":"cordis_inspect_self","arguments":{"pluginId":"snap-1"},"isError":false,"content":[{"type":"text","text":"{\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n}"}]}} -{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"advanced-code"},"content":[{"type":"tool-result","toolCallId":"advanced-code","content":[{"type":"text","text":"{\n \"run\": {\n \"status\": \"running\",\n \"pluginId\": \"snap-1\",\n \"packageId\": \"pkg-1\",\n \"pluginRunId\": \"run-1\",\n \"currentPackageId\": \"pkg-1\",\n \"host\": {\n \"status\": \"running\",\n \"provides\": [],\n \"waitingFor\": []\n },\n \"client\": {\n \"status\": \"absent\",\n \"waitingFor\": []\n }\n },\n \"inspected\": {\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n }\n}"}],"isError":false}],"role":"user","id":"0d0af8e0-841c-416f-9be6-bcbed9406669"}},"sourceEventSeqs":[24],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"advanced-code"},"content":[{"type":"tool-result","toolCallId":"advanced-code","content":[{"type":"text","text":"{\n \"run\": {\n \"status\": \"running\",\n \"pluginId\": \"snap-1\",\n \"packageId\": \"pkg-1\",\n \"pluginRunId\": \"run-1\",\n \"currentPackageId\": \"pkg-1\",\n \"host\": {\n \"status\": \"running\",\n \"provides\": [],\n \"waitingFor\": []\n },\n \"client\": {\n \"status\": \"absent\",\n \"waitingFor\": []\n }\n },\n \"inspected\": {\n \"mode\": \"plugin\",\n \"pluginId\": \"snap-1\",\n \"name\": \"Snapshot Marker\",\n \"packageCount\": 1,\n \"state\": \"running\",\n \"currentPackageId\": \"pkg-1\",\n \"activeRun\": {\n \"pluginRunId\": \"run-1\",\n \"packageId\": \"pkg-1\"\n },\n \"packages\": [\n {\n \"packageId\": \"pkg-1\",\n \"name\": \"Snapshot Marker\",\n \"purpose\": \"Exercise the dynamic Cordis Package lifecycle in the snapshot.\",\n \"hasHostHalf\": true,\n \"hasClientHalf\": false,\n \"isCurrent\": true,\n \"isNext\": false\n }\n ]\n }\n}"}],"isError":false}],"role":"user","id":"be73aaef-5172-4350-b03b-2fb5390f48ab"}},"sourceEventSeqs":[24],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"step/start","data":{"turn":1,"step":3}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -36,9 +36,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e1775d38-b17e-4e98-8dff-700d604cddbf"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"c506169d-8d5c-418d-a356-5824b0959032"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":3,"callId":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\",\"run_in_background\":false}"}} -{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"advanced-direct-child"},"content":[{"type":"tool-result","toolCallId":"advanced-direct-child","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"isError":false}],"role":"user","id":"a9ae23a0-f3b1-432e-ade5-0a25aec04838"}},"sourceEventSeqs":[38],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"advanced-direct-child"},"content":[{"type":"tool-result","toolCallId":"advanced-direct-child","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"isError":false}],"role":"user","id":"d057d856-694b-483c-a28a-bcbc4e8d829f"}},"sourceEventSeqs":[38],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":3}} {"type":"step/start","data":{"turn":1,"step":4}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -46,13 +46,13 @@ {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"3be25390-4ab3-4d98-a7b3-6b7df9c89f47"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"5dc26fbe-bb01-42a5-beb9-1188ac9338d1"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":4,"callId":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-headless-snapshot\",\"description\":\"exercise one workflow child through the headless agent\"}}"}} {"type":"tool-workflow/run-start","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","name":"advanced-headless-snapshot"}} {"type":"tool-workflow/agent-start","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","seq":1,"label":"workflow-child","phase":"Delegate","childId":"33333333-3333-4333-8333-333333333333"}} {"type":"tool-workflow/agent-end","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","seq":1,"outcome":"completed"}} {"type":"tool-workflow/run-end","data":{"runId":"cd3d2666-94a1-4285-804e-c99630bc7b51","stopReason":"completed"}} -{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"advanced-workflow"},"content":[{"type":"tool-result","toolCallId":"advanced-workflow","content":[{"type":"text","text":"workflow \"advanced-headless-snapshot\" completed (1 agent).\nReturn value:\n{\n \"reply\": \"WORKFLOW_CHILD_OK\"\n}"}],"isError":false}],"role":"user","id":"4d88debc-419d-4bea-ace5-e4626475cf83"}},"sourceEventSeqs":[48],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"advanced-workflow"},"content":[{"type":"tool-result","toolCallId":"advanced-workflow","content":[{"type":"text","text":"workflow \"advanced-headless-snapshot\" completed (1 agent).\nReturn value:\n{\n \"reply\": \"WORKFLOW_CHILD_OK\"\n}"}],"isError":false}],"role":"user","id":"d999a81b-5c7d-4103-a087-ae8c2ddb0ed5"}},"sourceEventSeqs":[48],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":4}} {"type":"step/start","data":{"turn":1,"step":5}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -60,9 +60,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"00552a01-9bbc-4fc0-82b5-a00140d5b12a"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[56,57,58,59,60],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"05fe05dc-6412-4d89-a285-c0134ac6c0ac"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[56,57,58,59,60],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":5,"callId":"advanced-undefine","name":"cordis_undefine","arguments":"{\"pluginId\":\"snap-1\"}"}} -{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"advanced-undefine"},"content":[{"type":"tool-result","toolCallId":"advanced-undefine","content":[{"type":"text","text":"Removed dynamic Plugin snap-1 and all of its Packages."}],"isError":false}],"role":"user","id":"af8f561e-1a9a-48e9-a94b-80e361fe5142"}},"sourceEventSeqs":[62],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"advanced-undefine"},"content":[{"type":"tool-result","toolCallId":"advanced-undefine","content":[{"type":"text","text":"Removed dynamic Plugin snap-1 and all of its Packages."}],"isError":false}],"role":"user","id":"da99fd98-3e31-4438-82ad-8cd5e8629181"}},"sourceEventSeqs":[62],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":5}} {"type":"step/start","data":{"turn":1,"step":6}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} @@ -70,6 +70,6 @@ {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ADVANCED_HEADLESS_OK"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"text","text":"ADVANCED_HEADLESS_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4556caeb-f947-40b9-a3c4-98e0f48b2edc"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[66,67,68,69,70],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"text","text":"ADVANCED_HEADLESS_OK"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"ee6bf425-ed0f-42cd-863d-aa0ee6b312dd"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[66,67,68,69,70],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":6}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl b/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl index 5963280e2c..d5894130a2 100644 --- a/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl +++ b/examples/headless-agent/tests/snapshots/compaction-recovery/session.jsonl @@ -1,32 +1,32 @@ {"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"7e4059c6-45e3-4067-8966-b2d402b97879"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"37b5d5fe-6796-4292-8c33-d60bdee54e72"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"7e4059c6-45e3-4067-8966-b2d402b97879"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Establish a durable compaction premise before continuing. Record every part of this historical evidence: the snapshot uses keyless replay; persistence uses JSONL; the assembled headless application loads its real Cordis composition; model-visible inputs remain logged; tool calls and results remain paired and ordered; context overflow retains the original failure while recovery is attempted; compaction opens with compaction/start and closes with compaction/end; a successful auxiliary summary records compaction/summary provenance; the replacement surface shadows only an older balanced range; the checkpoint remains smaller than the history it replaces; the newest tool result remains verbatim; the retried request sees that checkpoint; the final response proves the same turn continued; deterministic snapshot evidence stays separate from the live-provider smoke; and no external API key is needed. Emit one alpha marker through bash, then finish the task after any required recovery with the exact words COMPACTION RECOVERED."}],"source":{"kind":"user"},"role":"user","id":"37b5d5fe-6796-4292-8c33-d60bdee54e72"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Establish a durable compaction premise","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash","contextWindow":128000}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_compaction_marker","name":"bash","argumentsDelta":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":24,"outputTokens":6}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d1e07317-e842-4a27-bd7d-b679adebb42a"},"usage":{"inputTokens":24,"outputTokens":6}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4a9531dc-6e00-4321-878d-120fc32d6f50"},"usage":{"inputTokens":24,"outputTokens":6}},"sourceEventSeqs":[8,9,10,11,12],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"call_compaction_marker","name":"bash","arguments":"{\"command\":\"printf 'alpha\\n'\",\"description\":\"Emit compaction premise marker\"}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_compaction_marker"},"content":[{"type":"tool-result","toolCallId":"call_compaction_marker","content":[{"type":"text","text":"alpha\n"}],"isError":false}],"role":"user","id":"ed26c0aa-4a0a-49e3-88b4-795f48e02e45"}},"sourceEventSeqs":[14],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_compaction_marker"},"content":[{"type":"tool-result","toolCallId":"call_compaction_marker","content":[{"type":"text","text":"alpha\n"}],"isError":false}],"role":"user","id":"79ad618b-1f8e-4469-bf55-bf1d4e451520"}},"sourceEventSeqs":[14],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"error","failure":{"message":"snapshot request exceeded the model context window","code":"CONTEXT_WINDOW_EXCEEDED"}}}}} {"type":"compaction/start","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","turn":1}} {"type":"compaction/summary","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","summary":[{"type":"text","text":"The request established a durable compaction premise."}],"rawOutput":[{"type":"text","text":"The request established a durable compaction premise."}],"llmStreamCall":true,"shadowedRange":{"start":4,"end":4},"shadowedSeqs":[4],"shadowedTokenCount":266,"provider":"deepseek-official","model":"deepseek-v4-flash","maxTokens":32,"usage":{"inputTokens":20,"outputTokens":4}}} -{"type":"user/message","data":{"content":[{"type":"text","text":"This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.\n\n"},{"type":"text","text":"The request established a durable compaction premise."},{"type":"text","text":""}],"source":{"kind":"plugin","plugin":"compact","compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1"},"role":"user","id":"6f0985ae-7ff0-4923-80ab-0548d2cf466f"},"sourceEventSeqs":[19,20,4],"surfaceOp":{"op":"replace","start":4,"end":4}} +{"type":"user/message","data":{"content":[{"type":"text","text":"This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.\n\n"},{"type":"text","text":"The request established a durable compaction premise."},{"type":"text","text":""}],"source":{"kind":"plugin","plugin":"compact","compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1"},"role":"user","id":"4c8dec06-da41-4588-b99a-970997d3ec09"},"sourceEventSeqs":[19,20,4],"surfaceOp":{"op":"replace","start":4,"end":4}} {"type":"compaction/end","data":{"compactionId":"338e88fa-e78b-4d49-bd38-8f919e85f5e1","turn":1}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":0,"text":"COMPACTION RECOVERED"}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"COMPACTION RECOVERED"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":20,"outputTokens":4}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"text","text":"COMPACTION RECOVERED"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"e5a0b0a1-8f90-450c-b994-27eb0d084be0"},"usage":{"inputTokens":20,"outputTokens":4}},"sourceEventSeqs":[23,24,25,26,27],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"text","text":"COMPACTION RECOVERED"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"bcf13f76-b43e-4d0f-9a1c-1449f83ca779"},"usage":{"inputTokens":20,"outputTokens":4}},"sourceEventSeqs":[23,24,25,26,27],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl b/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl index b285d23511..a598ac7ad9 100644 --- a/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl +++ b/examples/headless-agent/tests/snapshots/pty-tools/session.jsonl @@ -1,21 +1,21 @@ {"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}","delegationDepth":0} -{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d20ef69c-ac94-4aed-9c4a-821a6ebff160"}]}} +{"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"73304b49-3a7c-4d92-ba69-3f34f4ebf101"}]}} {"type":"turn/start","data":{"turn":1}} {"type":"agent/inbox/spliced","data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}} {"type":"step/start","data":{"turn":1,"step":1}} -{"type":"user/message","data":{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"d20ef69c-ac94-4aed-9c4a-821a6ebff160"},"surfaceOp":"append"} -{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}]},"role":"user","id":"8731a232-5605-410b-9dd4-f7ae7b1a3f04"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Exercise the six PTY tools in order, including one missing-session signal error, then reply DONE."}],"source":{"kind":"user"},"role":"user","id":"73304b49-3a7c-4d92-ba69-3f34f4ebf101"},"surfaceOp":"append"} +{"type":"user/message","data":{"content":[{"type":"text","text":"Current runtime context. This snapshot supersedes earlier runtime-context snapshots.\n\nCurrent DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}],"source":{"kind":"plugin","plugin":"@deepseek-ai/dsh-system-prompt","form":"snapshot","sections":[{"name":"sandbox:policy","text":"Current DSH file policy: danger-full-access. The DSH file sandbox does not restrict file modifications by available operations."}]},"role":"user","id":"f411ee08-42d5-41bc-99b5-a33e0070f3c7"},"surfaceOp":"append"} {"type":"session/title","data":{"title":"Exercise the six PTY tools","messageSeqs":[4],"source":{"kind":"fallback"}}} -{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nUse a terminal session only when work needs persistent terminal state or interactive stdin; prefer shell/read/write/edit for bounded one-shot operations. Track every terminal session id and close sessions that no longer matter. An inferred_idle or timeout result does not prove the foreground command exited.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"terminal_close","description":"Close one persistent terminal and wait until its captured owned process tree is gone.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."}},"required":["sessionId"]}},{"name":"terminal_list","description":"List persistent terminal sessions owned by the current agent.","parameters":{"type":"object","properties":{}}},{"name":"terminal_open","description":"Create a persistent, owner-isolated terminal session from a registered backend type. Use this for shell or REPL state that must survive across tool calls.","parameters":{"type":"object","properties":{"type":{"type":"string","description":"Registered terminal backend type, usually \"shell\"."},"name":{"type":"string","description":"Optional owner-local display name such as \"main\" or \"gdb\"."},"cwd":{"type":"string","description":"Initial working directory. Defaults to the deployment workspace root."}},"required":["type"]}},{"name":"terminal_read","description":"Read a bounded page of retained output from a persistent terminal without sending input.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"offset":{"type":"number","description":"Newest-relative line offset (default 0)."},"count":{"type":"number","description":"Requested line count (default 500; backend caps apply)."}},"required":["sessionId"]}},{"name":"terminal_send","description":"Send text to a persistent terminal. By default Enter is submitted and the call waits for a prompt, stdin wait, output silence, timeout, or session exit. Background mode returns a job id for job_output/job_kill.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id returned by terminal_open or terminal_list."},"text":{"type":"string","description":"UTF-8 text to write to the terminal."},"submit":{"type":"boolean","description":"Submit Enter after text (default true). Set false for control characters or incomplete REPL input."},"run_in_background":{"type":"boolean","description":"Return a job id immediately; collect with job_output or stop with job_kill."}},"required":["sessionId","text"]}},{"name":"terminal_signal","description":"Send an allowed signal to the current foreground process group of a persistent terminal.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"signal":{"type":"string","description":"Signal to deliver. Shell-targeted SIGKILL is rejected; use terminal_close.","enum":["SIGINT","SIGTERM","SIGKILL","SIGTSTP","SIGHUP"]}},"required":["sessionId","signal"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} +{"type":"request/header","data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-flash"},"system":"You are an AI agent powered by DeepSeek Harness.\n\nYou are headless-agent, a coding assistant powered by the deepseek-v4-flash model.\n\nVerify your work by running the code or tests. Keep answers brief and\nfactual.\n\n\nUse the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.\n\nUse the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-observation-policy requires it) and prefer edit for targeted changes.\n\nUse the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-observation-policy requires it), unless you just created or edited it in this session.\n\nCheck the [exit code: N] marker on every bash result; investigate failures before moving on.\n\nUse a terminal session only when work needs persistent terminal state or interactive stdin; prefer shell/read/write/edit for bounded one-shot operations. Track every terminal session id and close sessions that no longer matter. An inferred_idle or timeout result does not prove the foreground command exited.\n\nTrack every background job id you start. You are notified in-session when a job finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running job's work. Before giving a final answer, collect every still-relevant job with job_output (set wait: true only when you are genuinely blocked on it), and job_kill jobs that stopped mattering.\n\nUse the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.\n\nUse the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.\n\nUse subagent in the background by default. Start independent delegations together in one assistant message and continue useful work while they run. Set `run_in_background: false` only when your next action depends on that subagent's result. When a background run settles, the runtime sends you a notice containing its outcome and any final assistant message.","tools":[{"name":"bash","description":"Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a job id immediately; read its output with `job_output` and stop it with `job_kill`.","parameters":{"type":"object","properties":{"command":{"type":"string","description":"The bash command to execute."},"description":{"type":"string","description":"Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."},"timeoutMs":{"type":"number","description":"Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."},"workdir":{"type":"string","description":"Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."},"run_in_background":{"type":"boolean","description":"Run in the background and return a job id immediately (collect with job_output, stop with job_kill). No timeout applies."}},"required":["command","description"]}},{"name":"edit","description":"Edit an existing UTF-8 text file by replacing literal text.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to edit, resolved by the filesystem backend."},"old_string":{"type":"string","description":"Literal text to replace. Must match exactly."},"new_string":{"type":"string","description":"Literal replacement text. Use an empty string to delete the match."},"replace_all":{"type":"boolean","description":"Replace all matches. Defaults to false; when false, old_string must appear exactly once."}},"required":["file_path","old_string","new_string"]}},{"name":"interrupt_agent","description":"Request cancellation of a background agent's current turn by its agent id. The target may be your direct child or a deeper agent created under you. Only the current turn stops: messages already queued for the agent stay parked until a later send_message, agents it started keep running, and the agent itself stays available for follow-ups. This call returns as soon as the stop request is accepted, so the target may keep running briefly; interrupting an agent that already finished is an accepted no-op.","parameters":{"type":"object","properties":{"agent_id":{"type":"string","description":"The agent id of the running agent to interrupt."}},"required":["agent_id"]}},{"name":"job_kill","description":"Request cancellation of a running background job by job id. Returns immediately; the job settles as killed once its work actually stops.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"reason":{"type":"string","description":"Optional short reason, recorded in the log and forwarded to the job."}},"required":["job_id"]}},{"name":"job_list","description":"List your background jobs (running and finished) with their ids, kinds, and statuses.","parameters":{"type":"object","properties":{}}},{"name":"job_output","description":"Read a background job. Stream jobs return only output since the previous read; final-output jobs return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.","parameters":{"type":"object","properties":{"job_id":{"type":"string","description":"Job id returned by the tool that started the background work."},"wait":{"type":"boolean","description":"Block until the job reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the job alive."},"timeout_ms":{"type":"number","description":"Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."}},"required":["job_id"]}},{"name":"ralph","description":"Run a foreground fresh-agent Ralph loop toward one immutable objective. Use only when the direct human explicitly asks for Ralph or fresh-agent iteration. Each round opens a new child with no parent conversation or prior child session; the shared workspace is long-term memory, and only a bounded structured report crosses rounds. The call returns when a worker reports completion or a concrete blocker, or at the round limit. Ordinary long-running same-session work belongs to goal tools.","parameters":{"type":"object","properties":{"objective":{"type":"string","description":"The immutable completion objective for every fresh Ralph round."},"maxRounds":{"type":"number","description":"Optional positive safe-integer round cap, bounded by the deployment ceiling."}},"required":["objective"]}},{"name":"read","description":"Read a UTF-8 text file and return line-numbered content.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to read, resolved by the filesystem backend."},"offset":{"type":"number","description":"1-based first line to return. Defaults to 1."},"limit":{"type":"number","description":"Maximum number of lines to return. Defaults to 2000."}},"required":["file_path"]}},{"name":"send_message","description":"Send a message to a background subagent by its subagent id, continuing the same conversation. It becomes the subagent's next turn: if it is still working, the message waits until its current turn finishes, so it cannot redirect work already underway. This call returns no answer from the subagent — only confirmation that the message was delivered — so use it to give it more work. A failure means the message was NOT delivered.","parameters":{"type":"object","properties":{"subagent_id":{"type":"string","description":"The subagent id returned when the background subagent was started."},"message":{"type":"string","description":"The message to deliver to the subagent."}},"required":["subagent_id","message"]}},{"name":"skill","description":"Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.","parameters":{"type":"object","properties":{"name":{"type":"string","description":"The exact skill name from the available skills list."}},"required":["name"]}},{"name":"subagent","description":"Delegate a self-contained task to a subagent (a separate agent that works in its own context) to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent returns its result, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. This tool runs in the background by default, immediately returns a durable subagent id, and keeps the child conversation available for later turns. When that run settles, the runtime sends the parent a notice containing its outcome and any final assistant message; `send_message` starts a later turn in the same child conversation. Set `run_in_background: false` only when your next action depends on receiving the result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."},"run_in_background":{"type":"boolean","description":"Whether to run in the background and return a durable subagent id immediately. Defaults to true. Set false to wait for the result when your next action depends on it."}},"required":["description","prompt"]}},{"name":"subagent_fork","description":"Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn). Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive its result, not its intermediate steps. This call waits for the subagent and returns its result.","parameters":{"type":"object","properties":{"description":{"type":"string","description":"A short (3-5 word) description of the delegated task, for display."},"prompt":{"type":"string","description":"The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."}},"required":["description","prompt"]}},{"name":"terminal_close","description":"Close one persistent terminal and wait until its captured owned process tree is gone.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."}},"required":["sessionId"]}},{"name":"terminal_list","description":"List persistent terminal sessions owned by the current agent.","parameters":{"type":"object","properties":{}}},{"name":"terminal_open","description":"Create a persistent, owner-isolated terminal session from a registered backend type. Use this for shell or REPL state that must survive across tool calls.","parameters":{"type":"object","properties":{"type":{"type":"string","description":"Registered terminal backend type, usually \"shell\"."},"name":{"type":"string","description":"Optional owner-local display name such as \"main\" or \"gdb\"."},"cwd":{"type":"string","description":"Initial working directory. Defaults to the deployment workspace root."}},"required":["type"]}},{"name":"terminal_read","description":"Read a bounded page of retained output from a persistent terminal without sending input.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"offset":{"type":"number","description":"Newest-relative line offset (default 0)."},"count":{"type":"number","description":"Requested line count (default 500; backend caps apply)."}},"required":["sessionId"]}},{"name":"terminal_send","description":"Send text to a persistent terminal. By default Enter is submitted and the call waits for a prompt, stdin wait, output silence, timeout, or session exit. Background mode returns a job id for job_output/job_kill.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id returned by terminal_open or terminal_list."},"text":{"type":"string","description":"UTF-8 text to write to the terminal."},"submit":{"type":"boolean","description":"Submit Enter after text (default true). Set false for control characters or incomplete REPL input."},"run_in_background":{"type":"boolean","description":"Return a job id immediately; collect with job_output or stop with job_kill."}},"required":["sessionId","text"]}},{"name":"terminal_signal","description":"Send an allowed signal to the current foreground process group of a persistent terminal.","parameters":{"type":"object","properties":{"sessionId":{"type":"string","description":"Terminal session id."},"signal":{"type":"string","description":"Signal to deliver. Shell-targeted SIGKILL is rejected; use terminal_close.","enum":["SIGINT","SIGTERM","SIGKILL","SIGTSTP","SIGHUP"]}},"required":["sessionId","signal"]}},{"name":"todo_write","description":"Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Mark every todo being actively worked on `in_progress` — several at once when work genuinely runs in parallel (e.g. concurrent subagents or background commands), one for sequential work; while work remains, at least one task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).","parameters":{"type":"object","properties":{"todos":{"type":"array","description":"The COMPLETE task list, replacing any previous list.","items":{"type":"object","additionalProperties":false,"properties":{"content":{"type":"string","description":"What the task is — a short imperative line."},"status":{"type":"string","description":"pending (not started) | in_progress (now) | completed (done).","enum":["pending","in_progress","completed"]}},"required":["content","status"]}}},"required":["todos"]}},{"name":"workflow","description":"Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.","parameters":{"type":"object","properties":{"script":{"type":"string","description":"The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return `)."},"meta":{"type":"object","description":"The workflow identity block (plain JSON — never code).","additionalProperties":true,"properties":{"name":{"type":"string","description":"Short kebab-case workflow name."},"description":{"type":"string","description":"One-line description of what the workflow does."},"whenToUse":{"type":"string","description":"Optional guidance on when this workflow applies."},"phases":{"type":"array","description":"Optional phase declarations matched by phase() calls.","items":{"type":"object","additionalProperties":true,"properties":{"title":{"type":"string","description":"The phase title phase() calls match by exact string."},"detail":{"type":"string","description":"Optional one-line description of the phase."},"provider":{"type":"string","description":"Optional provider override this phase is expected to use."},"model":{"type":"string","description":"Optional model override this phase is expected to use."}},"required":["title"]}}},"required":["name","description"]},"args":{"type":"object","description":"Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).","additionalProperties":true}},"required":["script","meta"]}},{"name":"write","description":"Create or fully replace a UTF-8 text file.","parameters":{"type":"object","properties":{"file_path":{"type":"string","description":"Path to write, resolved by the filesystem backend."},"content":{"type":"string","description":"Full UTF-8 text content to write."}},"required":["file_path","content"]}}]},"reason":"initial"}} {"type":"request/context","data":{"provider":"deepseek-official","model":"deepseek-v4-flash"}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"pty-spawn","name":"terminal_open","argumentsDelta":"{\"type\":\"shell\",\"name\":\"main\"}"}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d642809d-44ee-42cb-867d-63acb3d88b7a"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4fee9a25-1085-4f76-852d-7d4134da4771"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":1,"callId":"pty-spawn","name":"terminal_open","arguments":"{\"type\":\"shell\",\"name\":\"main\"}"}} -{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"pty-spawn"},"content":[{"type":"tool-result","toolCallId":"pty-spawn","content":[{"type":"text","text":"started terminal session pty-1 (main) [type: shell]\ndsh> "}],"isError":false}],"role":"user","id":"f5d9b371-8400-4993-9cac-f61fcc1aab9b"}},"sourceEventSeqs":[15],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"pty-spawn"},"content":[{"type":"tool-result","toolCallId":"pty-spawn","content":[{"type":"text","text":"started terminal session pty-1 (main) [type: shell]\ndsh> "}],"isError":false}],"role":"user","id":"b2d286b5-c84e-4368-91f5-0ca31a26fcc8"}},"sourceEventSeqs":[15],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":1}} {"type":"step/start","data":{"turn":1,"step":2}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -23,9 +23,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"bf54676e-46ad-41b6-8938-c5b5cf9429ea"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"0f63c607-e2c1-4004-af07-05621b5cbffa"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":2,"callId":"pty-send","name":"terminal_send","arguments":"{\"sessionId\":\"pty-1\",\"text\":\"printf 'PTY_OK\\\\n'\"}"}} -{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"pty-send"},"content":[{"type":"tool-result","toolCallId":"pty-send","content":[{"type":"text","text":"K\ndsh> \n[wait: stdin_read]\n[session: running]\n[output truncated]"}],"isError":false}],"role":"user","id":"1ea99fa1-54bd-45cd-8935-041298013a91"},"meta":{"viewport":"printf 'PTY_OK\\n'\nPTY_OK\ndsh> ","waitReason":"stdin_read","sessionStatus":{"kind":"running"},"truncated":false}},"sourceEventSeqs":[25],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"pty-send"},"content":[{"type":"tool-result","toolCallId":"pty-send","content":[{"type":"text","text":"K\ndsh> \n[wait: stdin_read]\n[session: running]\n[output truncated]"}],"isError":false}],"role":"user","id":"41fcd151-f242-4946-b94f-ff1b2a813f24"},"meta":{"viewport":"printf 'PTY_OK\\n'\nPTY_OK\ndsh> ","waitReason":"stdin_read","sessionStatus":{"kind":"running"},"truncated":false}},"sourceEventSeqs":[25],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":2}} {"type":"step/start","data":{"turn":1,"step":3}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -33,9 +33,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"cb394133-a39a-4c7c-bba0-f51a06a1ddd9"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[29,30,31,32,33],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"0b3f3d05-bec4-4c05-9d10-def19db565ca"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[29,30,31,32,33],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":3,"callId":"pty-read","name":"terminal_read","arguments":"{\"sessionId\":\"pty-1\",\"offset\":0,\"count\":20}"}} -{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"pty-read"},"content":[{"type":"tool-result","toolCallId":"pty-read","content":[{"type":"text","text":"dsh> printf 'PTY_OK\\n'\nPTY_OK\ndsh> \n[lines: 0-3 of 3]"}],"isError":false}],"role":"user","id":"116cdcf0-42dc-4220-8f01-38ed1153f903"}},"sourceEventSeqs":[35],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"pty-read"},"content":[{"type":"tool-result","toolCallId":"pty-read","content":[{"type":"text","text":"dsh> printf 'PTY_OK\\n'\nPTY_OK\ndsh> \n[lines: 0-3 of 3]"}],"isError":false}],"role":"user","id":"21e52dde-c420-408d-996b-d695d8cbcb5e"}},"sourceEventSeqs":[35],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":3}} {"type":"step/start","data":{"turn":1,"step":4}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -43,9 +43,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b7216958-bc3a-4631-bf30-31db4f96b5f4"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[39,40,41,42,43],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"4591bab3-687f-4bcf-b236-105f6650f603"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[39,40,41,42,43],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":4,"callId":"pty-signal","name":"terminal_signal","arguments":"{\"sessionId\":\"pty-missing\",\"signal\":\"SIGINT\"}"}} -{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"pty-signal"},"content":[{"type":"tool-result","toolCallId":"pty-signal","content":[{"type":"text","text":"Error: unknown PTY session pty-missing"}],"isError":true}],"role":"user","id":"d950a25e-05ca-4eec-aaf8-9937b8fe4f48"}},"sourceEventSeqs":[45],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"pty-signal"},"content":[{"type":"tool-result","toolCallId":"pty-signal","content":[{"type":"text","text":"Error: unknown PTY session pty-missing"}],"isError":true}],"role":"user","id":"641383e7-6af3-4067-84eb-a3f9e17488cb"}},"sourceEventSeqs":[45],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":4}} {"type":"step/start","data":{"turn":1,"step":5}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -53,9 +53,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"6dae8392-a011-47e5-84d8-d9e4c3c19d0f"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"36cf385b-b4e9-4e99-831d-cd8def844f75"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":5,"callId":"pty-kill","name":"terminal_close","arguments":"{\"sessionId\":\"pty-1\"}"}} -{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"pty-kill"},"content":[{"type":"tool-result","toolCallId":"pty-kill","content":[{"type":"text","text":"closed terminal session pty-1"}],"isError":false}],"role":"user","id":"fb3d7081-eba7-4169-a2fe-e05ecae357ac"}},"sourceEventSeqs":[55],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"pty-kill"},"content":[{"type":"tool-result","toolCallId":"pty-kill","content":[{"type":"text","text":"closed terminal session pty-1"}],"isError":false}],"role":"user","id":"14f60758-1455-4ecb-a001-6775d61c6886"}},"sourceEventSeqs":[55],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":5}} {"type":"step/start","data":{"turn":1,"step":6}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} @@ -63,9 +63,9 @@ {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}} {"type":"assistant/chunk","data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} -{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b4305101-fbda-4a57-ac80-1c6cdae233f9"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[59,60,61,62,63],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"tool-call","id":"pty-list","name":"terminal_list","arguments":"{}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"d00ae522-a67d-4c01-b83c-c31093efb7ba"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[59,60,61,62,63],"surfaceOp":"append"} {"type":"tool/call","data":{"turn":1,"step":6,"callId":"pty-list","name":"terminal_list","arguments":"{}"}} -{"type":"tool/result","data":{"turn":1,"step":6,"message":{"source":{"kind":"tool","callId":"pty-list"},"content":[{"type":"tool-result","toolCallId":"pty-list","content":[{"type":"text","text":"(no terminal sessions)"}],"isError":false}],"role":"user","id":"0097f12a-a04e-4042-b47a-4a1535708049"}},"sourceEventSeqs":[65],"surfaceOp":"append"} +{"type":"tool/result","data":{"turn":1,"step":6,"message":{"source":{"kind":"tool","callId":"pty-list"},"content":[{"type":"tool-result","toolCallId":"pty-list","content":[{"type":"text","text":"(no terminal sessions)"}],"isError":false}],"role":"user","id":"efbe4f79-cb4b-43e0-979b-b9310b5f2691"}},"sourceEventSeqs":[65],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":6}} {"type":"step/start","data":{"turn":1,"step":7}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} @@ -73,6 +73,6 @@ {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":3}}}} {"type":"assistant/chunk","data":{"turn":1,"step":7,"chunk":{"type":"finish","reason":{"kind":"stop"}}}} -{"type":"assistant/message","data":{"turn":1,"step":7,"message":{"role":"assistant","content":[{"type":"text","text":"DONE"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"b75bf882-4d9d-4f92-9125-0cdc18f76f25"},"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[69,70,71,72,73],"surfaceOp":"append"} +{"type":"assistant/message","data":{"turn":1,"step":7,"message":{"role":"assistant","content":[{"type":"text","text":"DONE"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-flash"},"id":"fdeff700-5f7b-4264-9ada-47b846a74354"},"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[69,70,71,72,73],"surfaceOp":"append"} {"type":"step/end","data":{"turn":1,"step":7}} {"type":"turn/end","data":{"turn":1,"reason":{"kind":"completed"}}} diff --git a/packages/workflow/tool-ralph/src/index.ts b/packages/workflow/tool-ralph/src/index.ts index d92d7251e2..3e3e1b5943 100644 --- a/packages/workflow/tool-ralph/src/index.ts +++ b/packages/workflow/tool-ralph/src/index.ts @@ -406,7 +406,7 @@ export function apply(ctx: Context, config: Config): void { const resolved = resolveConfig(config) ctx.systemPrompt.section({ name: 'tool:ralph', - order: 117, + order: 116, text: 'Use the ralph tool ONLY when the direct human explicitly asks for a Ralph loop or fresh-agent iterative execution. Each Ralph round starts a fresh child with no conversation seed and uses the shared workspace as durable memory. Completion and blockers are worker reports, not independent evaluation. Use same-session goal tools for ordinary long-running objectives, and plain subagents or workflows for bounded delegation and fan-out.', }) ctx.tools.register(defineTool({ diff --git a/packages/workflow/tool-workflow/src/index.ts b/packages/workflow/tool-workflow/src/index.ts index 899c6f8ab7..7128fe4a95 100644 --- a/packages/workflow/tool-workflow/src/index.ts +++ b/packages/workflow/tool-workflow/src/index.ts @@ -211,7 +211,7 @@ export function apply(ctx: Context, config: Config): void { // lives in tool plugins as prompt sections, not in the deployment persona). ctx.systemPrompt.section({ name: `tool:${toolName}`, - order: 116, + order: 115.5, text: `Use the ${toolName} tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.`, }) ctx.tools.register(defineTool({