Merge remote-tracking branch 'origin/worktree/deepseek-native-multimodal' into worktree/deepseek-vision-model-catalog

.agents/notes/implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.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/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.md
-2026-07-10-single-file-executable-sdk-runtime-distribution.md: fa2f86893b730aa1ba020bd568d268ec8d9d6239
-2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: 509bec18edb9923dd4d60d4ecf30d4fbcd9cc6d5
+2026-07-10-single-file-executable-sdk-runtime-distribution.md: 40433d99e5d1aa569c3fdf094a280d3de62ad588
+2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: 54030fa4b0742fbc282bc327b0ca22747e6a20bd

.agents/notes/implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.md

@@ -36,19 +36,21 @@ Config discovery has two channels and fails loudly when both are missing: the `D
 
 Inside the exe's VFS sits a **real package tree in build-artifact form** (each package's `lib/` plus a real `node_modules`). The packaged JSON-RPC entry supplies its installed harness base to app-boot's root Include: relative plugin specifiers resolve from the external configuration directory, while bare package names resolve from the VFS, so a configuration inside another Node project cannot shadow the packaged plugin set. The ordinary development bin leaves bare packages configuration-owned. Bare specifiers in the packaged entry resolve upward along `node_modules` from the entry's position inside the VFS and land inside the VFS naturally. The closed set needs no allowlist code — the set is whatever the VFS has installed, and importing a name outside the set fails.
 
-The deploy root is [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json) (`dsh-jsonrpc-agent-pkg`, a pnpm workspace member and a zero-code pure dependency manifest) — the unified source of truth for "which plugins the exe ships" and "what the Python runtime distributes". Adding a plugin to the exe = adding one dependency line to the manifest and repackaging. [`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) traverses every workspace package covered by that manifest and requires every non-optional workspace peer at the runtime root, reporting the complete referencing-package → missing-peer chain; `pnpm run hygiene`, CI static, and the single-exe build run it before packaging. Deploy also packs by each package's `files`, so the shared chunks tsdown splits out must be covered by `files`.
+The deploy root is [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json) (`dsh-jsonrpc-agent-pkg`, a pnpm workspace member and a zero-code pure dependency manifest) — the unified source of truth for "which plugins the exe ships" and "what the Python runtime distributes". Adding a plugin to the exe = adding one dependency line to the manifest and repackaging. [`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) reads every shipped `apps/cli/config/agent-presets/*/agent.cordis.yml`, evaluates `disabled` conditions that compare `process.platform` for every target in `python/sdk-runtime/platforms.json`, and requires each active workspace plugin at the runtime root through an explicit `workspace:` dependency. It also traverses every workspace package covered by that manifest and requires every non-optional workspace peer, reporting the complete preset or referencing-package → missing-dependency chain; unknown platform conditions remain active so a plugin cannot be omitted by an unsupported expression. `pnpm run hygiene`, CI static, and the single-exe build run it before packaging. Deploy also packs by each package's `files`, so the shared chunks tsdown splits out must be covered by `files`.
+
+The deploy root includes `@deepseek-ai/dsh-mcp-client` as an explicitly supported custom-configuration plugin even though no shipped preset mounts it. An external config can therefore connect to user-supplied stdio and Streamable HTTP MCP servers and register their tools; the distribution does not carry those servers or extend the bridge to MCP Resources and Prompts. The executable and installed-wheel smokes start a temporary stdio server, discover its tool, and complete one model-requested call.
 
 ### Build pipeline and artifacts
 
-[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → restore any direct workspace package that legacy deploy hoisted back under the source manifest's `node_modules`, omitting its package-local dependency tree and rejecting any remaining manifest gap → replace every staged dependency symlink with its target bytes, remove package-manager `.bin` links, and fail if any symlink remains → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source; CI rebuilds that addon inside the matching manylinux 2.28 container before packaging, and the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory. macOS uses its target prebuild and emits the required `-spawn-helper` beside the executable. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted gives pkg a stable single-instance layout that the explicit materialization pass makes symlink-free; disabling automatic peer installation prevents undeclared peers from expanding the closure; link-workspace-packages selects direct workspace dependencies. [`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) overrides the transitive `@deepseek-ai/cosmokit` and `@deepseek-ai/schemastery` semver requests to the pinned vendor sources so legacy deploy never resolves those unpublished names from a registry.
+[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → restore any direct workspace package that legacy deploy hoisted back under the source manifest's `node_modules`, omitting its package-local dependency tree and rejecting any remaining manifest gap → replace every staged dependency symlink with its target bytes, remove package-manager `.bin` links, and fail if any symlink remains → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source; CI rebuilds that addon inside the matching manylinux 2.28 container before packaging, and the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory. Every target copies its native `@vscode/ripgrep` binary beside the executable as the required `-rg` sidecar; pkg runtimes select that sidecar through `process.pkg`, while ordinary Node execution uses `@vscode/ripgrep` directly. macOS uses its target prebuild and also emits the required `-spawn-helper`. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted gives pkg a stable single-instance layout that the explicit materialization pass makes symlink-free; disabling automatic peer installation prevents undeclared peers from expanding the closure; link-workspace-packages selects direct workspace dependencies. [`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) overrides the transitive `@deepseek-ai/cosmokit` and `@deepseek-ai/schemastery` semver requests to the pinned vendor sources so legacy deploy never resolves those unpublished names from a registry.
 
 CI: [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml), called for linux-x64 by the [required Python runtime pull-request validation](../testing/2026-08-12-required-python-runtime-pull-request-ci.md), triggered explicitly by `workflow_dispatch` or the `build-exe` label for selected targets, and called for all targets by the [public publication workflow](../process/2026-08-11-python-publication-workflow.md). Native builds run on linux-x64 / linux-arm64 (`ubuntu-24.04-arm`) / macos-arm64, with `~/.pkg-cache` cached, and pkg handles macOS ad-hoc signing. Each leg drives a mock SSE model through the SDK with the default config and a custom `cordis.yml`, drives the exe directly over NDJSON JSON-RPC, verifies the JSONL and final response, and installs release-shaped wheels into a clean venv without `runtime_bin`; Linux additionally inspects both the executable and native addon's GLIBC requirements and runs in a manylinux 2.28 container, while macOS verifies that the executable's deployment target fits the wheel tag. A full three-target run retains four artifacts, each containing one release file: the platform-independent SDK wheel and three native runtime wheels; a subset dispatch retains the SDK wheel and selected runtime wheels. Bare executables and source bundles remain intermediate test inputs. [`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) accepts `python-v<repository-version>` tag pipelines whose version matches the root `package.json`, builds one SDK wheel and three native runtime wheels, then a single serialized job checks and publishes all four to the project PyPI registry. Windows is a non-goal.
 
 ### Python SDK distribution: two carriers, exe for production, node for development
 
-The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe and optional helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.
+The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe with its required `-rg` sidecar and optional macOS helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.
 
-[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) reads the authoritative `X.Y.Z` or prerelease version from the repository root `package.json`, converts prereleases to their PEP 440 spelling, and stages both packages at that wheel version, with `deepseek-harness-sdk` depending exactly on the matching `deepseek-harness-runtime-bin`. An optional `python-v<repository-version>` release tag is a consistency assertion and is rejected when it differs from the repository version; the source `pyproject.toml` development sentinel never determines a release version. Staging also carries the repository license into both wheels and the third-party notices into the bundled runtime wheel. The SDK is a `py3-none-any` wheel; each wheel-only runtime package contains one exe, and the macOS wheel also contains its architecture-matched helper. Runtime wheels use one of `py3-none-manylinux_2_28_x86_64`, `py3-none-manylinux_2_28_aarch64`, or the conservative `py3-none-macosx_14_0_arm64` tag for the Node 24 executable's macOS 13.5 deployment target; the Hatch hook rejects sdists, universal tags, mixed-platform payloads, missing or extra helpers, and unsupported platforms.
+[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) reads the authoritative `X.Y.Z` or prerelease version from the repository root `package.json`, converts prereleases to their PEP 440 spelling, and stages both packages at that wheel version, with `deepseek-harness-sdk` depending exactly on the matching `deepseek-harness-runtime-bin`. An optional `python-v<repository-version>` release tag is a consistency assertion and is rejected when it differs from the repository version; the source `pyproject.toml` development sentinel never determines a release version. Staging also carries the repository license into both wheels and the third-party notices into the bundled runtime wheel. The SDK is a `py3-none-any` wheel; each wheel-only runtime package contains one exe and its architecture-matched `-rg` sidecar, and the macOS wheel also contains its architecture-matched spawn helper. Runtime wheels use one of `py3-none-manylinux_2_28_x86_64`, `py3-none-manylinux_2_28_aarch64`, or the conservative `py3-none-macosx_14_0_arm64` tag for the Node 24 executable's macOS 13.5 deployment target; the Hatch hook rejects sdists, universal tags, mixed-platform payloads, missing or extra sidecars, and unsupported platforms.
 
 The exe's "must be explicitly configured" hard semantic is unchanged; the zero-config experience is restored by the wrapper: when the caller gave no `cordis`, named no explicit runtime, and the environment has no `DSH_CORDIS_CONFIG`, the client explicitly injects the checked-in default `cordis.yml` (agent-core + preloaded llm-deepseek + JSONL persistence + bash-local + the `dsh-sdk-jsonrpc-server` serving entry, with `!!js` environment-variable fallbacks) via `DSH_CORDIS_CONFIG`.
 
@@ -62,7 +64,7 @@ The exe's "must be explicitly configured" hard semantic is unchanged; the zero-c
 
 ## Testing
 
-The verification surface has three tiers. Mechanism tier: the measured conclusions for the `--sea` chain are embedded in the Decision sections (ESM dynamic import inside the VFS, single cordis instance, fail-loud config chain, `node:sqlite`, macOS ad-hoc signing runs). SDK tier: the complete keyless pytest suite covers the client protocol against a fake runtime peer, subprocess cleanup, absolute cwd propagation, dual-carrier launch, and carrier resolution; root CI runs it on Python 3.10. End-to-end tier: every platform build completes a turn against a mock endpoint through the default SDK path, a custom config, the checked-in standalone minimal composition, and the direct binary protocol, with final text and JSONL checked. The minimal run asserts its exact system prompt and two-tool catalog, retains Bash state across calls, and invokes the editor. The custom config additionally drives `run_code` and a zero-agent `workflow` through their real worker files inside the packaged VFS. The same build leg runs a committed executable-specific snapshot through the Python SDK: a keyless scripted model mounts a Cordis plugin that registers a tool, invokes that tool from `run_code`, runs a direct spawn subagent and a workflow that starts a second spawn child, then unmounts the plugin. The fixture explicitly disables its unused bundled Bash and local skill discovery so its tool set does not depend on repository-external state, and the comparison normalizes opaque message, agent, workflow-run, and session IDs across the SDK result and notification stream plus the parent and two child JSONL logs. This harness stays separate from ACP's `pnpm run test:snapshot` because the protocols and build artifacts differ. The platform wheel is then installed in a clean venv and run without `runtime_bin`.
+The verification surface has three tiers. Mechanism tier: the measured conclusions for the `--sea` chain are embedded in the Decision sections (ESM dynamic import inside the VFS, single cordis instance, fail-loud config chain, `node:sqlite`, macOS ad-hoc signing runs). SDK tier: the complete keyless pytest suite covers the client protocol against a fake runtime peer, subprocess cleanup, absolute cwd propagation, dual-carrier launch, and carrier resolution; root CI runs it on Python 3.10. End-to-end tier: every platform build completes a turn against a mock endpoint through the default SDK path, a custom config, the checked-in standalone minimal composition, and the direct binary protocol, with final text and JSONL checked. The minimal run asserts its exact system prompt and two-tool catalog, retains Bash state across calls, and invokes the editor. The custom config additionally drives `run_code` and a zero-agent `workflow` through their real worker files inside the packaged VFS. The filesystem-search scenario requires the model to call both `glob` and `grep` through the target-native `-rg` sidecar. The MCP scenario starts a temporary external stdio server, deliberately delays its initial `tools/list` response, then immediately starts the first SDK prompt; the prompt must see and call the discovered tool, proving that `initialize` is a real Loader-settlement readiness boundary rather than a timing sleep. The same build leg runs a committed executable-specific snapshot through the Python SDK: a keyless scripted model mounts a Cordis plugin that registers a tool, invokes that tool from `run_code`, runs a direct spawn subagent and a workflow that starts a second spawn child, then unmounts the plugin. The fixture explicitly disables its unused bundled Bash and local skill discovery so its tool set does not depend on repository-external state, and the comparison normalizes opaque message, agent, workflow-run, and session IDs across the SDK result and notification stream plus the parent and two child JSONL logs. This harness stays separate from ACP's `pnpm run test:snapshot` because the protocols and build artifacts differ. The platform wheel is then installed in a clean venv and run without `runtime_bin`.
 
 Manual-driving caveat: the bin treats stdin EOF as "the client is gone" and disposes immediately, so a short-lived pipe aborts an in-flight turn — pipe-driven runs must keep stdin open until the turn ends.
 

.agents/notes/implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md

@@ -36,19 +36,21 @@ exe 使用 [@yao-pkg/pkg](https://github.com/yao-pkg/pkg)（vercel/pkg 归档后
 
 exe 的 VFS 内是**构建产物形态的真实包树**（各包的 `lib/` + 真实 `node_modules`）。打包专用 JSON-RPC 入口会向 app-boot 的根 Include 提供自身已安装 harness 的基准位置：相对插件说明符从外部配置目录解析，裸包名则从 VFS 解析，因此位于另一个 Node 项目内的配置无法遮蔽已打包的插件集合。普通开发 bin 仍由配置项目提供裸包。打包入口中的裸包名从该入口在 VFS 内的位置沿 `node_modules` 向上解析，自然落在 VFS 内。封闭集不需要白名单代码——VFS 中安装了什么，集合中就有什么；`import()` 集合外的名称会失败。
 
-部署根目录是 [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json)（`dsh-jsonrpc-agent-pkg`，pnpm 工作区成员、零代码纯依赖 manifest），也是「exe 安装哪些插件」与「Python 运行时分发什么」的统一真源。向 exe 添加插件，就是在 manifest 中增加一行依赖后重新打包。[`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) 遍历该 manifest 覆盖的全部工作区包，要求每个非可选的工作区对等依赖（peer dependency）都显式列在运行时根目录，并报告“引用包 → 缺失对等依赖”的完整链路；`pnpm run hygiene`、CI 静态检查与 single-exe 构建都会在打包前运行该门禁。部署还会依据各包的 `files` 字段打包，因此 tsdown 拆出的共享分片必须被 `files` 覆盖。
+部署根目录是 [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json)（`dsh-jsonrpc-agent-pkg`，pnpm 工作区成员、零代码纯依赖 manifest），也是「exe 安装哪些插件」与「Python 运行时分发什么」的统一真源。向 exe 添加插件，就是在 manifest 中增加一行依赖后重新打包。[`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) 读取每个已发布的 `apps/cli/config/agent-presets/*/agent.cordis.yml`，针对 `python/sdk-runtime/platforms.json` 中的每个目标解析比较 `process.platform` 的 `disabled` 条件，并要求该目标启用的每个工作区插件都通过显式的 `workspace:` 依赖列在运行时根目录。它还遍历该 manifest 覆盖的全部工作区包，要求每个非可选的工作区对等依赖（peer dependency）都显式列出，并报告“preset 或引用包 → 缺失依赖”的完整链路；无法识别的平台条件会保持启用，避免因不支持的表达式遗漏插件。`pnpm run hygiene`、CI 静态检查与 single-exe 构建都会在打包前运行该门禁。部署还会依据各包的 `files` 字段打包，因此 tsdown 拆出的共享分片必须被 `files` 覆盖。
+
+部署根目录显式包含 `@deepseek-ai/dsh-mcp-client`，将其作为自定义配置可用的插件，即使随附 preset 均未挂载该插件。外部配置因此可以连接由用户提供的 stdio 与 Streamable HTTP MCP server 并注册其工具；分发物不包含这些 server，也不将桥接范围扩展到 MCP Resources 和 Prompts。可执行程序与已安装 wheel 包的冒烟测试会启动临时 stdio server，发现其工具，并完成一次由模型请求的调用。
 
 ### 构建流水线与产物
 
-[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 恢复被 legacy deploy 提升回源 manifest 的 `node_modules` 下的任何直接工作区包，同时省略其包内依赖树，并拒绝剩余的 manifest 缺口 → 将暂存依赖中的每个符号链接替换为目标文件内容，删除包管理器的 `.bin` 链接，并在仍有任何符号链接时失败 → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`；CI 会在打包前进入匹配架构的 manylinux 2.28 容器重新构建该 addon，而 `--legacy` 部署会省略这一副作用目录，因此构建器会把它从根安装目录复制到暂存闭包。macOS 使用对应目标的预构建产物，并在可执行文件旁生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 为 pkg 提供稳定的单实例布局，再由显式物化步骤消除符号链接；关闭对等依赖自动安装可防止未声明的对等依赖扩大闭包；`link-workspace-packages` 选择直接工作区依赖。[`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) 将传递的 `@deepseek-ai/cosmokit` 与 `@deepseek-ai/schemastery` semver 请求覆盖到固定的 vendor 源码，使 legacy deploy 不会从注册表解析这些未发布名称。
+[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 恢复被 legacy deploy 提升回源 manifest 的 `node_modules` 下的任何直接工作区包，同时省略其包内依赖树，并拒绝剩余的 manifest 缺口 → 将暂存依赖中的每个符号链接替换为目标文件内容，删除包管理器的 `.bin` 链接，并在仍有任何符号链接时失败 → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`；CI 会在打包前进入匹配架构的 manylinux 2.28 容器重新构建该 addon，而 `--legacy` 部署会省略这一副作用目录，因此构建器会把它从根安装目录复制到暂存闭包。每个目标都会把对应的原生 `@vscode/ripgrep` 二进制复制到可执行文件旁，作为必需的 `-rg` 伴随文件；pkg 运行时通过 `process.pkg` 选择该伴随文件，普通 Node 执行则直接使用 `@vscode/ripgrep`。macOS 使用对应目标的预构建产物，并额外生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 为 pkg 提供稳定的单实例布局，再由显式物化步骤消除符号链接；关闭对等依赖自动安装可防止未声明的对等依赖扩大闭包；`link-workspace-packages` 选择直接工作区依赖。[`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) 将传递的 `@deepseek-ai/cosmokit` 与 `@deepseek-ai/schemastery` semver 请求覆盖到固定的 vendor 源码，使 legacy deploy 不会从注册表解析这些未发布名称。
 
 CI 使用 [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml)：[必需的 Python 运行时拉取请求验证](../testing/2026-08-12-required-python-runtime-pull-request-ci.md)调用它构建 linux-x64，手动派发 `workflow_dispatch` 或 PR（Pull Request）的 `build-exe` 标签可以显式选择构建目标，[公开发布工作流](../process/2026-08-11-python-publication-workflow.md)则调用它构建全部目标。linux-x64、linux-arm64（`ubuntu-24.04-arm`）和 macos-arm64 三个平台分别进行原生构建，并缓存 `~/.pkg-cache`；macOS 的 ad-hoc 签名由 pkg 处理。每个平台都使用 mock SSE（Server-Sent Events）模型，分别通过默认配置和自定义 `cordis.yml` 驱动 SDK，再通过 NDJSON JSON-RPC 直接驱动 exe，校验 JSONL 与最终响应；最后把发布形态的 wheel 包安装到干净的 venv 中，并在不传 `runtime_bin` 的情况下运行。Linux 还会检查可执行文件和原生 addon 各自的 GLIBC 依赖，并在 manylinux 2.28 容器中运行；macOS 则验证可执行文件的部署目标符合 wheel 包标签。完整构建三个目标时保留 4 个产物，每个产物只含一个发布文件：平台无关的 SDK wheel 包与 3 个原生运行时 wheel 包；手动选择部分目标时保留 SDK wheel 与所选运行时 wheel。裸 exe 与源码包只作为测试中间输入。[`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) 只接受版本与根目录 `package.json` 匹配的 `python-v<repository-version>` 标签流水线，构建一个 SDK wheel 包和 3 个原生运行时 wheel 包，再由单个串行任务校验并将这 4 个文件发布到项目的 PyPI 注册表。Windows 不在目标范围内。
 
 ### Python SDK 分发：双载体，exe 用于生产，`node` 用于开发
 
-Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 与可选 helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。
+Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 及其必需的 `-rg` 伴随文件和可选的 macOS helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。
 
-[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) 从仓库根目录的 `package.json` 读取权威的 `X.Y.Z` 或预发布版本，把预发布版本转换为 PEP 440 写法，并以该 wheel 包版本暂存两个包，让 `deepseek-harness-sdk` 精确依赖匹配版本的 `deepseek-harness-runtime-bin`。可选的 `python-v<repository-version>` 发布标签只是一项一致性断言，与仓库版本不同时会被拒绝；源码 `pyproject.toml` 中的开发占位版本从不决定发布版本。暂存过程还会把仓库许可证放入两个 wheel 包，并把第三方声明放入内置运行时 wheel 包。SDK 是 `py3-none-any` wheel 包；每个只提供 wheel 包的运行时包都包含一个 exe，macOS wheel 包还包含与其架构匹配的 helper。运行时 wheel 包使用 `py3-none-manylinux_2_28_x86_64`、`py3-none-manylinux_2_28_aarch64`，或针对 Node 24 可执行文件 macOS 13.5 部署目标而保守选择的 `py3-none-macosx_14_0_arm64` 标签；Hatch 钩子拒绝 sdist、通用标签、混合平台载荷、helper 缺失或多余，以及不支持的平台。
+[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) 从仓库根目录的 `package.json` 读取权威的 `X.Y.Z` 或预发布版本，把预发布版本转换为 PEP 440 写法，并以该 wheel 包版本暂存两个包，让 `deepseek-harness-sdk` 精确依赖匹配版本的 `deepseek-harness-runtime-bin`。可选的 `python-v<repository-version>` 发布标签只是一项一致性断言，与仓库版本不同时会被拒绝；源码 `pyproject.toml` 中的开发占位版本从不决定发布版本。暂存过程还会把仓库许可证放入两个 wheel 包，并把第三方声明放入内置运行时 wheel 包。SDK 是 `py3-none-any` wheel 包；每个只提供 wheel 包的运行时包都包含一个 exe 及其架构匹配的 `-rg` 伴随文件，macOS wheel 包还包含与其架构匹配的 spawn helper。运行时 wheel 包使用 `py3-none-manylinux_2_28_x86_64`、`py3-none-manylinux_2_28_aarch64`，或针对 Node 24 可执行文件 macOS 13.5 部署目标而保守选择的 `py3-none-macosx_14_0_arm64` 标签；Hatch 钩子拒绝 sdist、通用标签、混合平台载荷、伴随文件缺失或多余，以及不支持的平台。
 
 exe「必须显式配置」的硬语义不变；零配置体验由包装层恢复：调用方没有提供 `cordis`、没有显式指定运行时，且环境中没有 `DSH_CORDIS_CONFIG` 时，客户端将检入的默认 `cordis.yml`（`agent-core` + 预载的 `llm-deepseek` + JSONL 持久化 + `bash-local` + `dsh-sdk-jsonrpc-server` 对外服务条目，并通过 `!!js` 使用环境变量兜底）显式注入 `DSH_CORDIS_CONFIG`。
 
@@ -62,7 +64,7 @@ exe 内支持 `dsh-workflow-worker-thread` 与 `dsh-code-runtime-worker-thread`
 
 ## 测试
 
-验证面分三层。机制层：`--sea` 链路的实测结论内嵌在「决策」各节（VFS 内 ESM 动态 `import()`、单一 Cordis 实例、明确报错的配置链路、`node:sqlite`、macOS ad-hoc 签名可运行）。SDK 层：完整的无密钥 pytest 套件以 mock 运行时对端覆盖客户端协议、子进程清理、绝对 `cwd` 传递、双载体启动与载体解析；根 CI 在 Python 3.10 上运行全部用例。端到端层：每个平台构建都通过默认 SDK 路径、自定义配置、仓库内置的独立 minimal 组合和直接二进制协议，对 mock 端点完成一个轮次，并校验最终文本与 JSONL。minimal 运行会断言其精确系统提示词与双工具目录，跨调用保留 Bash 状态，并调用编辑器。自定义配置还会通过打包进 VFS 的真实工作线程文件执行 `run_code` 和不启动 agent 的 `workflow`。同一构建任务还会经 Python SDK 运行一组检入的 exe 专用快照：无密钥脚本化模型挂载一个会注册工具的 Cordis 插件，从 `run_code` 调用该工具，运行一个直接 spawn 的 subagent 和一个会通过 spawn 启动第二个 subagent 的工作流，随后卸载该插件。该 fixture（测试前置数据）会显式禁用组合包中未使用的 Bash 和本地 skill（技能）发现，使其工具集不依赖仓库外部状态；比较时会规范化 SDK 结果与通知流，以及父会话和两个子会话 JSONL 日志中不透明的消息、agent、工作流运行与会话 ID。该 harness 与 ACP 的 `pnpm run test:snapshot` 保持独立，因为二者的协议和构建产物不同。随后把平台 wheel 包安装进干净的 venv，并在不传 `runtime_bin` 的情况下运行。
+验证面分三层。机制层：`--sea` 链路的实测结论内嵌在「决策」各节（VFS 内 ESM 动态 `import()`、单一 Cordis 实例、明确报错的配置链路、`node:sqlite`、macOS ad-hoc 签名可运行）。SDK 层：完整的无密钥 pytest 套件以 mock 运行时对端覆盖客户端协议、子进程清理、绝对 `cwd` 传递、双载体启动与载体解析；根 CI 在 Python 3.10 上运行全部用例。端到端层：每个平台构建都通过默认 SDK 路径、自定义配置、仓库内置的独立 minimal 组合和直接二进制协议，对 mock 端点完成一个轮次，并校验最终文本与 JSONL。minimal 运行会断言其精确系统提示词与双工具目录，跨调用保留 Bash 状态，并调用编辑器。自定义配置还会通过打包进 VFS 的真实工作线程文件执行 `run_code` 和不启动 agent 的 `workflow`。文件系统搜索场景要求模型通过目标平台的 `-rg` 伴随文件调用 `glob` 与 `grep`。MCP 场景会启动临时外部 stdio server，刻意延迟首次 `tools/list` 响应，随后立即启动第一个 SDK 提示词；该提示词必须看到并调用已发现的工具，从而证明 `initialize` 是真正以 Loader 插件树完全稳定为准的就绪边界，而不是依赖定时 sleep。同一构建任务还会经 Python SDK 运行一组检入的 exe 专用快照：无密钥脚本化模型挂载一个会注册工具的 Cordis 插件，从 `run_code` 调用该工具，运行一个直接 spawn 的 subagent 和一个会通过 spawn 启动第二个 subagent 的工作流，随后卸载该插件。该 fixture（测试前置数据）会显式禁用组合包中未使用的 Bash 和本地 skill（技能）发现，使其工具集不依赖仓库外部状态；比较时会规范化 SDK 结果与通知流，以及父会话和两个子会话 JSONL 日志中不透明的消息、agent、工作流运行与会话 ID。该 harness 与 ACP 的 `pnpm run test:snapshot` 保持独立，因为二者的协议和构建产物不同。随后把平台 wheel 包安装进干净的 venv，并在不传 `runtime_bin` 的情况下运行。
 
 
 手工驱动注意：`bin` 将 stdin EOF 视为「客户端已离开」并立即 dispose，生命周期较短的管道会中止进行中的轮次——管道驱动必须保持 stdin 打开，直到轮次结束。

.agents/notes/implemented/bug-fix/2026-08-13-feedback-note-editor-popover.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/2026-08-13-feedback-note-editor-popover.md
+2026-08-13-feedback-note-editor-popover.md: 33b7cd84b97ceac7fef1d96f1dee279fbb215800
+2026-08-13-feedback-note-editor-popover.zh.md: b130ab1e9c66372cf2a52bc5e12f65027c64cf51

.agents/notes/implemented/bug-fix/2026-08-13-feedback-note-editor-popover.md

@@ -0,0 +1,43 @@
+# Agent Note: The feedback note editor floats above the transcript in a popover
+
+Status: implemented
+
+English | [中文](2026-08-13-feedback-note-editor-popover.zh.md)
+
+## Problem
+
+The Web surface for message feedback ([#2262](https://github.com/deepseek-harness/deepseek-harness/pull/2262)) contributes its controls to `conversation.chat.assistant-actions`, which renders inside the finalized assistant message's shared IconActions row. That row was one fixed-height `flex` line with `flex-wrap` at its initial `nowrap` and `height: 28px`, sized for 28px icons and a clock. The note editor mounted into it as an inline group holding a `width: 260px` textarea plus Save and Cancel.
+
+A 260px input and two buttons do not fit that line at any window size. Measured against the shipped bundle, the row's scrollable overflow with the editor open was 168px at a 1680px viewport and 444px at 600px — the defect was never a narrow-window edge case, it was present at full-screen desktop. Flex overflow spills past the end of the line, so the items after the editor in flex order were the ones pushed out of the conversation column: the branch action left the column at 600px, and the clock and its run/TTFT/throughput readings left it at 900px. Those controls stay hit-testable while invisible, so no behavioral assertion noticed; the shipped e2e covered rate, note, reload, and retract, and the 24 UI snapshots are width-independent DOM.
+
+The same stylesheet also named four `--dsw-alias-*` tokens that the theme does not define: `border-secondary`, `bg-primary`, `interactive-bg-primary`, and `label-inverse`. An undefined custom property makes its whole declaration invalid at computed-value time, so the textarea shipped with no border and no surface, and Save with neither fill nor a readable label — the editor read as loose text floating in the transcript rather than as an input.
+
+## Decision
+
+The note editor does not enter the row's flex layout at all. It is a popover: a fixed-position panel, portaled to `document.body`, whose coordinates come from the note trigger's rect. The row keeps its single line of icons and the note trigger, so nothing has to shrink, wrap, or reflow around the editor, and no `order` or wrapping is needed anywhere. Portaling out of the conversation column also escapes its `overflow` clip, so the panel cannot be cropped at the scroll edge and it moves with the message it annotates when the transcript scrolls. This reuses the same portal mechanism `ui-primitives/Menu` uses for anchored menus (`ui-subagent`'s catalog popover is built on it): the panel is `position: fixed`, placed from the anchor rect on open, clamped inside the viewport, and re-placed on scroll (capture phase) and resize. That anchoring is shared rather than copied: `ui-primitives/useAnchoredPosition` owns measure-offset-clamp-and-track, and the duplication gate is what forced the extraction — an inline copy of the clamp and its listener pair reported a 10-line clone against `Menu`. `Menu` keeps its own effect because its placement also resolves `side`/`align` variants and an optional caller-supplied anchor rect, which this surface does not need; the hook covers the plain below-the-anchor case both would otherwise spell out.
+
+**The action strip.** The like/dislike buttons and the note trigger stay in the row, unchanged. The trigger is a plain button (`aria-haspopup="dialog"`, `aria-expanded` while open) that shows "Add a note" before a note exists and the note text afterward.
+
+**The popover.** While open, the panel contains the textarea plus Save and Cancel, and any note-save failure, as `role="dialog"` with a title distinct from the textarea's own label so both are addressable by name. It opens beneath the trigger (4px gap), clamps to 12px from the viewport edges, auto-focuses the textarea, and closes on Escape or an outside pointer-down. Closing returns focus to the trigger only when the panel was really open, never on the initial mount (a freshly rendered rated message must not pull focus into its action row). A rating action during an open editor closes the panel. The four undefined tokens are replaced with the ones the theme actually defines, matching the primitives' precedent: `border-l2` and `bg-layer-1` for the input, `button-primary-fill` with `label-primary-foreground` plus a `button-primary-hover` state for Save; the panel surface reuses the Menu card recipe (`--dsw-specific-menu`, `--dsw-shadow-lv3`, inverted hairline `--dsw-alias-border-inverted`, `border-radius: 12px`).
+
+**Failure surfaces split by where the human is looking.** A rating or list-load failure shows beside the buttons in the row, legible whether or not the popover is open. A note-save failure shows inside the popover, next to Save/Cancel, and the panel stays open so the draft survives to be corrected.
+
+## Alternatives considered
+
+**Inline expansion on the row, the editor claiming its own line via a full-width flex basis with the row allowed to wrap** — the approach first shipped on this branch and rejected here. It fixes the geometry (the row reports zero overflow from 1680px down to 600px) but at a visible cost: the branch action and the end clock wrap below the editor while it is open, the row occupies three lines, and the interaction competes for the same horizontal strip the row already fills. That cost is what [#2561](https://github.com/deepseek-harness/deepseek-harness/issues/2561) reported from real use — the row reads as misaligned once the editor expands — and it asked for the popover the chat surface already uses. A popover removes the editor from the row entirely, so the strip and the keyboard tab order are untouched whether the editor is open or not.
+
+**An absolutely-positioned popover not portaled out of the column** — rejected: the conversation column is an `overflow-y: auto` scroller, so a panel laid out inside it is clipped at the scroll edge and does not track the message as the column scrolls. Portaling to `document.body` with fixed placement from the trigger rect is what makes the floating panel viable, exactly as `Menu`'s portal mode and the subagent catalog popover already do.
+
+**A new `belowActions` seam on `MessageIconActions`, rendering the editor as a sibling under the row** — rejected: the slot contract documents `assistant-actions` as rendering *inside* the message's IconActions row, and one entry cannot supply two render sites without widening the host contract for a presentation detail that a portaled popover already expresses without touching the host.
+
+## Consequences
+
+With the editor open the actions row stays a single 28px line with zero overflow and nothing outside the column at every viewport from 1680px down to 600px — because the editor is not in the row to begin with. The panel floats above the transcript inside the viewport and stays anchored to its trigger, escaping the column's overflow clip. The editor is legible as an input in both themes.
+
+`apps/web/tests/message-feedback-layout.e2e.ts` sweeps six viewports with the editor open and pins, per stop, that the row reports one line and zero overflow, that the panel is outside the conversation column (proof it escapes the clip), that it lies within the viewport (proof the clamp holds), and that it sits by its trigger. A committed golden records the relations; reverting to inline (or dropping the portal) fails the geometry assertions. `packages/client/ui-message-feedback/tests/styles.client.spec.ts` checks the tokens against the theme's committed source, that the panel is `position: fixed`, and that it carries no flex sizing (so it cannot rejoin the row), plus the brace balance, following the `ui-settings-models` styles-spec precedent. The unit spec covers rate, note, reload, retract, plus the popover's portal-to-body, Escape/outside-click dismissal, and keep-open-on-inside-click.
+
+The `ui-message-feedback` package adds `@types/react-dom` so the `createPortal` usage typechecks, mirroring `ui-primitives`.
+
+Known limitations are accepted rather than fixed here. A rating click while the panel is open closes it, and the close path returns focus to the note trigger rather than leaving it on the rating button the human just pressed; the same happens when an outside click lands on another focusable control, which the browser focuses before the close returns focus to the trigger. A pointer user does not notice either; a keyboard user feels the focus move. The clamp assumes the panel fits: a panel taller than the viewport makes the upper bound `innerHeight - height - margin` smaller than `margin`, so `top` goes negative and the panel's head is cut off rather than its foot. The panel's three-row textarea carries `resize: vertical`, so a human can drag past that size; `.notePanel` therefore bounds its height at `calc(100vh - 24px)` and scrolls its own content, the counterpart of the existing `max-width` and the same 12px margin the clamp uses. If the rating disappears while the editor is open, the panel unmounts on the `rating !== undefined` guard but `noteOpen` stays true, so the document-level Escape and pointer-down listeners remain attached; should the item reappear through a later resync, the panel returns with the previous draft and without refocusing the textarea. The window is one click or Escape wide, and the save failure it could hide already falls back to the row, so it is left as it is. A failure that lands after the panel was closed and reopened is not written into the new session's panel: its draft was reseeded from the stored note, so an old attempt's error would mislabel it, and the uncommitted content is already gone — the failure is dropped rather than shown. And while the placement replays on scroll, window resize, and the panel's own size changes, jsdom has no layout, so the real geometry is proven by the browser scenario while the unit spec covers the wiring through a `ResizeObserver` stub.
+
+A residual narrow-viewport clock overflow remains below 520px from the clock string alone, unrelated to the feedback surface. The repo has no gate for undefined design tokens, and a scan during this work found more in `ui-agent-preset`, `ui-conversation`, `ui-jobs`, `ui-settings-plugins`, and `ui-tool`; they are untouched here and want their own change.

.agents/notes/implemented/bug-fix/2026-08-13-feedback-note-editor-popover.zh.md

@@ -0,0 +1,43 @@
+# Agent Note：反馈备注编辑器以浮层悬浮在对话记录上方
+
+Status: implemented
+
+[English](2026-08-13-feedback-note-editor-popover.md) | 中文
+
+## Problem
+
+消息反馈的 Web 界面（[#2262](https://github.com/deepseek-harness/deepseek-harness/pull/2262)）把控件贡献给 `conversation.chat.assistant-actions`，该槽位渲染在已定稿助手消息共享的 IconActions 行内。那一行是单条固定高度的 `flex` 线，`flex-wrap` 保持初始值 `nowrap` 且 `height: 28px`，按 28px 图标加一个时钟来定尺寸。备注编辑器作为一个内联组挂进去，内含 `width: 260px` 的 textarea 加 Save 与 Cancel。
+
+一个 260px 输入框加两个按钮在任何窗口尺寸下都装不进那条线。对着已构建产物实测，编辑器打开时该行的可滚动溢出在 1680px 视口下是 168px，在 600px 下是 444px——这个缺陷从来不是窄窗口的边缘情况，在全屏桌面下就已存在。flex 溢出会溢出到线的末端之外，因此按 flex 顺序排在编辑器之后的项被挤出会话列：branch 操作在 600px 时离开列，时钟及其运行时长/TTFT/吞吐读数在 900px 时离开列。这些控件在不可见的同时仍可命中测试，所以没有任何行为断言发现它；已交付的 e2e 覆盖评分、备注、reload 与撤回，而 24 个 UI 快照是与宽度无关的 DOM。
+
+同一张样式表还引用了四个主题并未定义的 `--dsw-alias-*` token：`border-secondary`、`bg-primary`、`interactive-bg-primary` 与 `label-inverse`。未定义的自定义属性会让其所在的整条声明在 computed-value 阶段失效，因此 textarea 交付时既无边框也无底色，Save 既无填充也无可读标签——编辑器读起来像是浮在对话记录里的散落文本，而不是一个输入框。
+
+## Decision
+
+备注编辑器完全不进入行的 flex 布局。它是一个浮层：一张固定定位的面板，portal 到 `document.body`，其坐标来自备注触发按钮的矩形。行保持其单行图标与备注触发按钮，因此没有任何东西需要围绕编辑器收缩、换行或回流，任何地方都不需要 `order` 或换行。portal 出会话列也逃出了列的 `overflow` 裁剪，因此面板不会被滚动边缘裁掉，并且当对话记录滚动时会随它所批注的消息一起移动。这里复用 `ui-primitives/Menu` 为锚定菜单所用的同一套 portal 机制（`ui-subagent` 的 catalog popover 就构建在它之上）：面板 `position: fixed`，打开时从 anchor rect 定位，钳制在视口内，并在滚动（捕获阶段）与缩放时重新定位。这套锚定逻辑是共享而非复制的：`ui-primitives/useAnchoredPosition` 持有「测量—偏移—钳制—跟随」这一件事，而促成这次抽取的正是重复代码门禁——内联的钳制与那对监听器被报为与 `Menu` 的 10 行克隆。`Menu` 保留自己的 effect，因为它的定位还要解析 `side`/`align` 变体与可选的调用方 anchor rect，而本界面不需要这些；该 hook 覆盖的是两边本来都要各写一遍的「锚点正下方」这一简单情形。
+
+**操作条。** 点赞/点踩按钮与备注触发按钮保持原样留在行内。触发按钮是普通 `button`（`aria-haspopup="dialog"`，打开时 `aria-expanded`），在没有备注时显示「补充说明」，已有备注时显示备注文本。
+
+**浮层。** 打开时，面板内含 textarea、Save 与 Cancel，以及任何备注保存失败提示，作为 `role="dialog"`，其标题与 textarea 自身的标签不同，以便两者都能按名称寻址。它在触发按钮下方打开（4px 间距），钳制到距视口边缘 12px，自动聚焦 textarea，并在 Escape 或外部 pointer-down 时关闭。关闭时仅当面板确实曾经打开才把焦点还给触发按钮，绝不会在初始挂载时（新渲染出的一条已评分消息不得把焦点拉进其操作条）。编辑器打开时进行评分操作会关闭面板。四个未定义 token 换成主题确实定义的那些，与 primitives 的既有做法一致：输入框用 `border-l2` 与 `bg-layer-1`，Save 用 `button-primary-fill` 配 `label-primary-foreground` 并加 `button-primary-hover` 状态；面板表面复用 Menu 卡片的配方（`--dsw-specific-menu`、`--dsw-shadow-lv3`、反色发丝线 `--dsw-alias-border-inverted`、`border-radius: 12px`）。
+
+**失败提示按人的视线所落之处拆分。** 评分或列表加载失败显示在按钮旁的图标行里，无论浮层是否打开都清晰可读。备注保存失败显示在浮层内、Save/Cancel 旁，且面板保持打开，以便草稿留存待修正。
+
+## Alternatives considered
+
+**行内展开：编辑器通过整行 flex basis 独占一行，并让行允许换行** — 这是本分支最初交付、在此否决的做法。它修好了几何（行在 1680px 到 600px 报告零溢出），但有可见代价：branch 与末尾时钟在编辑器打开时换行到编辑器下方，行占三行，交互与行本就占满的横向条带争空间。这一代价正是 [#2561](https://github.com/deepseek-harness/deepseek-harness/issues/2561) 在真实使用中反馈的问题——编辑器展开后这一行读起来是错位的——并提出改用 chat 界面已有的弹窗。浮层把编辑器完全移出行，因此无论编辑器是否打开，操作条与键盘 Tab 顺序都不受影响。
+
+**不 portal 出列的绝对定位浮层** — 否决：会话列是 `overflow-y: auto` 的滚动容器，因此在列内布局的面板会被滚动边缘裁掉，且不随列滚动而跟住消息。portal 到 `document.body` 并从触发按钮矩形做固定定位，才让浮动面板可行，正如 `Menu` 的 portal 模式与 subagent catalog popover 已然做到的那样。
+
+**在 `MessageIconActions` 上新增 `belowActions` 接缝，把编辑器作为该行的兄弟节点渲染在下方** — 否决：slot 契约明确记载 `assistant-actions` 渲染在消息 IconActions 行**内部**，且单个条目无法在不为一个展示细节拓宽 Host 契约的前提下提供两个渲染点，而 portal 出的浮层无需触碰 Host 就表达了该细节。
+
+## Consequences
+
+编辑器打开时，操作行保持单条 28px 线，在 1680px 到 600px 的每一档视口都零溢出、零项落在列外——因为编辑器本就不在行里。面板悬浮于对话记录之上、位于视口内，并保持锚定其触发按钮，逃出列溢出裁剪。编辑器在两种主题下都能被辨认为输入框。
+
+`apps/web/tests/message-feedback-layout.e2e.ts` 在编辑器打开时扫描六个视口，并在每一档钉住：行报告单行零溢出、面板位于会话列之外（证明它逃出裁剪）、面板落在视口内（证明钳制有效）、面板紧贴其触发按钮。已提交的 golden 记录这些关系；回退到行内（或去掉 portal）会让几何断言失败。`packages/client/ui-message-feedback/tests/styles.client.spec.ts` 校验 token 与主题已提交的源一致、面板为 `position: fixed`、且不带任何 flex sizing（因此不会重新加入行），并校验大括号平衡，沿用 `ui-settings-models` styles spec 的先例。单元 spec 覆盖评分、备注、reload、撤回，外加浮层的 portal 到 body、Escape/外部点击关闭、以及浮层内部点击保持打开。
+
+`ui-message-feedback` 包新增 `@types/react-dom`，使 `createPortal` 用法能通过类型检查，与 `ui-primitives` 一致。
+
+有若干已知限制在此接受而非修复。面板打开时点击评分会关闭它，而关闭路径把焦点归还给备注触发按钮，而不是留在用户刚按下的评分按钮上；外部点击落在另一个可聚焦控件上时同理——浏览器先把焦点给该控件，随后关闭路径又把它拉回触发按钮。指针用户对两者都无感，键盘用户会察觉焦点移动。钳制假定面板放得下：面板高于视口时，上界 `innerHeight - height - margin` 会小于 `margin`，于是 `top` 变为负值、被裁掉的是面板顶部而非底部。面板里的三行 textarea 带 `resize: vertical`，用户可以拖过这个尺寸，因此 `.notePanel` 把自身高度限制在 `calc(100vh - 24px)` 并自行滚动内容——这是既有 `max-width` 的对应项，用的是与钳制相同的 12px 边距。编辑器打开时若评分消失，面板会因 `rating !== undefined` 守卫卸载，但 `noteOpen` 仍为 true，因此 document 级的 Escape 与 pointer-down 监听继续挂着；若该 item 之后经 resync 重新出现，浮层会带着上一次的草稿回来且不重新聚焦 textarea。该窗口只有一次点击或一次 Escape 那么宽，而它可能遮住的保存失败已经有行内回退，因此保持现状。若失败在面板关闭并重开后才到达，不会写入新会话的面板：其草稿已按已存备注重新播种，旧尝试的错误会误标新草稿，而未提交的内容本就不存在——该失败被丢弃而不展示。以及，定位虽然会在滚动、窗口缩放与面板自身尺寸变化时重放，但 jsdom 没有布局，因此真实几何由浏览器场景证明，单测则通过 `ResizeObserver` stub 覆盖其接线。
+
+520px 以下仍残留仅来自时钟字符串的窄视口溢出，与本界面无关。仓库没有针对未定义设计 token 的门禁；本次工作中的一次扫描在 `ui-agent-preset`、`ui-conversation`、`ui-jobs`、`ui-settings-plugins` 与 `ui-tool` 中又发现更多，本次未触碰，需要单独的改动处理。

.agents/notes/implemented/bug-fix/2026-08-18-tool-row-file-open-failure.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/2026-08-18-tool-row-file-open-failure.md
+2026-08-18-tool-row-file-open-failure.md: e36552395b992e688fad35b3163b92c9f6189e43
+2026-08-18-tool-row-file-open-failure.zh.md: 72b6026d9eb44c74f13c987996e652008f35a78e

.agents/notes/implemented/bug-fix/2026-08-18-tool-row-file-open-failure.md

@@ -0,0 +1,33 @@
+# Agent Note: Tool-row file-open failures stay visible
+
+Status: implemented
+
+English | [中文](2026-08-18-tool-row-file-open-failure.zh.md)
+
+## Problem
+
+Tool-row path clicks already call `host.openPath` through the chat view's injected `openFile`. The inject swallowed every Host or OS refusal, so a missing desktop opener, a remote or non-loopback carrier, or a path the Host cannot hand off left the row looking successful. The reader had no reason and no second try.
+
+The [file-open-in-OS decision](../feature/2026-07-28-tool-call-file-open-in-os.md) still owns the link gesture and the Host handoff. This note owns only the refusal.
+
+## Decision
+
+The inject returns the `workspaces.openPath` promise. The chat view wraps that opener: a rejection opens an in-page Modal with the thrown text (or the unknown-open fallback when that text is empty) and a Retry that repeats the same path; Cancel, Escape, the close control, and a mask click dismiss it. A later settlement after dismiss is ignored, so a cancelled in-flight refusal cannot reopen the dialog.
+
+The dialog lives on the view that owns the Host call, not on each tool row. Produced-file chips and closing-message mentions use the same wrapper because they already share that opener. The produced-files folder action opens `.`, and that refusal uses the folder title and unknown-open copy.
+
+The Host message is shown as thrown. `WorkspaceRuntime.openPath` prefixes `path open failed: ` onto the wire error; the dialog does not unwrap that prefix.
+
+## Alternatives considered
+
+- **Per-row inline error.** The Host call is conversation-owned and several entries share one opener; a row-local banner would duplicate the same refusal next to every click target.
+- **Toast without retry.** The product ask is the reason *and* a retry entry. The workspace folder-adoption dialog already pairs those two.
+- **Chat-store remount persistence.** A failed open is transient view state. The chat store survives view remounts, so a leftover dialog would return after a tab switch that cannot usefully retry the original gesture.
+
+## Consequences
+
+A silent Host refusal is no longer a success from the reader's seat. Headless or remote deployments that click a path now see why the desktop handoff did not happen. The view holds one extra request-generation counter so dismiss and retry stay race-safe.
+
+## Testing
+
+Package specs cover inject rejection, the dialog copy (Error, non-Error, empty, workspace folder), retry of the same path, cancel, and a settlement that arrives after dismiss. `apps/web/tests/seeded-history.e2e.ts` stubs `host.openPath` to fail over a cold-resumed read row, pins the assembled dialog in `file-open-failure.expected.md`, and asserts the English reason plus a second call with the same payload.

.agents/notes/implemented/bug-fix/2026-08-18-tool-row-file-open-failure.zh.md

@@ -0,0 +1,33 @@
+# Agent Note: Tool-row file-open failures stay visible
+
+Status: implemented
+
+[English](2026-08-18-tool-row-file-open-failure.md) | 中文
+
+## 问题
+
+工具行路径点击已经通过聊天视图注入的 `openFile` 调用 `host.openPath`。inject 吞掉了每一次 Host 或操作系统拒绝，因此缺少桌面打开器、远程或非回环载体、或 Host 无法交接的路径，都会让该行看起来像成功。读者看不到原因，也无法再试一次。
+
+[用系统应用打开文件的决策](../feature/2026-07-28-tool-call-file-open-in-os.md) 仍然拥有链接手势和 Host 交接。本 Agent Note 只拥有拒绝路径。
+
+## 决策
+
+inject 返回 `workspaces.openPath` 的 promise。聊天视图包装该打开器：拒绝时打开页面内 Modal，展示抛出的文本（文本为空时用未知打开回退文案），并提供对同一路径的重试；取消、Escape、关闭控件和点击遮罩会关掉对话框。关闭之后才落到的结果会被忽略，因此已取消的进行中拒绝不能再次打开对话框。
+
+对话框位于 chat 视图（拥有 Host 调用），而不是每个工具行。产物文件标签和收尾消息中的提及已经共用该打开器，因此走同一包装。产物文件的文件夹操作打开 `.`，该拒绝使用文件夹标题和未知打开回退文案。
+
+Host 消息按抛出内容展示。`WorkspaceRuntime.openPath` 会在 wire 错误前加上 `path open failed: ` 前缀；对话框不拆掉该前缀。
+
+## 考虑过的替代方案
+
+- **按行内联错误。** Host 调用由会话拥有，多个入口共用一个打开器；行内横幅会在每个点击目标旁重复同一拒绝。
+- **没有重试的 toast。** 产品要求同时给出原因和重试入口。工作区文件夹采纳对话框已经把这两者配对。
+- **写入 chat store 并跨 remount 保留。** 打开失败是瞬时视图状态。chat store 会在视图 remount 后存活，于是残留对话框会在无法有效重试原手势的页签切换之后回来。
+
+## 后果
+
+从读者一侧看，静默的 Host 拒绝不再等同于成功。无头或远程部署点击路径时，能看到桌面交接为何没有发生。视图多持有一个请求世代计数器，使关闭与重试在竞态下仍然安全。
+
+## 测试
+
+包测试覆盖 inject 拒绝、对话框文案（Error、非 Error、空文本、工作区文件夹）、同一路径重试、取消，以及关闭之后才落到的结果。`apps/web/tests/seeded-history.e2e.ts` 在冷恢复的 read 行上把 `host.openPath` stub 为失败，用 `file-open-failure.expected.md` 钉住组装后的对话框，并断言英文原因以及对同一 payload 的第二次调用。

.agents/notes/implemented/feature/2026-07-28-tool-call-file-open-in-os.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/feature/2026-07-28-tool-call-file-open-in-os.md
-2026-07-28-tool-call-file-open-in-os.md: d78cf4da2c1861a66a0cefb24dba785c7cb1279d
-2026-07-28-tool-call-file-open-in-os.zh.md: c13f6e60c0c70a3036f50c678c7b348c91f874c3
+2026-07-28-tool-call-file-open-in-os.md: 08fc51cc3a5d2fb43b67dc158fd1ee789fafdb68
+2026-07-28-tool-call-file-open-in-os.zh.md: 59079533b502d234bafb0b53c123e5895e105833

.agents/notes/implemented/feature/2026-07-28-tool-call-file-open-in-os.md

@@ -23,9 +23,9 @@ File-tool path summaries (`read` / `write` / `edit` args carrying `path` or `fil
 
 ## Consequences
 
-Clicking a file path in a tool row opens that path on the host. Non-file tool rows are inert summaries (expand toggles remain where the row already supported them). Remote or non-loopback clients cannot invoke `host.openPath`.
+Clicking a file path in a tool row opens that path on the host. Non-file tool rows are inert summaries (expand toggles remain where the row already supported them). Remote or non-loopback clients cannot invoke `host.openPath`. A Host or OS refusal is owned by the chat view: it shows the thrown reason and retries the same path ([file-open failure](../bug-fix/2026-08-18-tool-row-file-open-failure.md)).
 
 ## Risks
 
-- Desktop Linux hosts without `xdg-open`, and WSL hosts without working Windows interop (`wslpath` plus `powershell.exe`), fail the RPC; the chat row stays silent while the host returns an internal error.
+- Desktop Linux hosts without `xdg-open`, and WSL hosts without working Windows interop (`wslpath` plus `powershell.exe`), fail the RPC; the chat view shows that Host error and offers retry.
 - Relative paths without a session cwd are forwarded verbatim and may fail on the host.

.agents/notes/implemented/feature/2026-07-28-tool-call-file-open-in-os.zh.md

@@ -23,9 +23,9 @@ Status: implemented
 
 ## 后果
 
-点击工具行中的文件路径会在宿主上打开该路径。非文件工具行只是不可交互的摘要（行内已有的展开开关仍保留）。远程或非回环客户端无法调用 `host.openPath`。
+点击工具行中的文件路径会在宿主上打开该路径。非文件工具行只是不可交互的摘要（行内已有的展开开关仍保留）。远程或非回环客户端无法调用 `host.openPath`。Host 或操作系统拒绝由聊天视图拥有：它展示抛出的原因，并对同一路径提供重试（[打开失败](../bug-fix/2026-08-18-tool-row-file-open-failure.md)）。
 
 ## 风险
 
-- 没有 `xdg-open` 的桌面 Linux 宿主，以及 Windows 互操作（`wslpath` 加 `powershell.exe`）不可用的 WSL 宿主，会使 RPC 失败；聊天行保持静默，宿主返回内部错误。
+- 没有 `xdg-open` 的桌面 Linux 宿主，以及 Windows 互操作（`wslpath` 加 `powershell.exe`）不可用的 WSL 宿主，会使 RPC 失败；聊天视图展示该 Host 错误并提供重试。
 - 没有会话 cwd 时相对路径会原样转发，可能在宿主侧失败。

.agents/notes/implemented/feature/2026-08-04-claude-code-and-codex-subagent-backends.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/feature/2026-08-04-claude-code-and-codex-subagent-backends.md
-2026-08-04-claude-code-and-codex-subagent-backends.md: a8f500c7fb934b8634456e1618498909f682f0e2
-2026-08-04-claude-code-and-codex-subagent-backends.zh.md: ca35c38617b1ca38757959735b11618559f6f804
+2026-08-04-claude-code-and-codex-subagent-backends.md: 9b47fcf49d47d2c3561245fa1e16ff8c5da0a35c
+2026-08-04-claude-code-and-codex-subagent-backends.zh.md: fcb1aac71be2da9d907ad67867c763e3051baec5

.agents/notes/implemented/feature/2026-08-04-claude-code-and-codex-subagent-backends.md

@@ -12,7 +12,7 @@ The product integrations must not become second owners for task text, cwd, cance
 
 ## Decision
 
-The harness publishes two sibling one-shot provider packages whose default registry names are `codex` and `claude-code`. This note owns their product protocols, result mapping, and process lifecycle; the [named-instance decision](2026-08-18-product-subagent-named-instances.md) owns Profile-selected provider identity and static tool binding, the [production-install exclusion decision](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md) owns their independent optional Bundles and host-plane placement, the [product one-shot background decision](2026-08-12-product-subagent-one-shot-background-tasks.md) owns the model-visible scheduling choice, and the [non-interactive permissions decision](2026-08-15-product-subagent-noninteractive-permissions.md) owns each product Provider's Profile-selected mode and diagnostic production. Both packages accept multiple named instances. Loading either provider starts no product process, and each tool accepts only a standalone text task; product and instance selection remain deployment configuration.
+The harness publishes two sibling one-shot provider packages whose default registry names are `codex` and `claude-code`. This note owns their product protocols, result mapping, and process lifecycle; the [named-instance decision](2026-08-18-product-subagent-named-instances.md) owns Profile-selected provider identity and static tool binding, the [production-install exclusion decision](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md) owns their independent optional Bundles and host-plane placement, the [product one-shot background decision](2026-08-12-product-subagent-one-shot-background-tasks.md) owns the model-visible scheduling choice, the [non-interactive permissions decision](2026-08-15-product-subagent-noninteractive-permissions.md) owns each product Provider's Profile-selected mode and safe permission decisions, and the [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns version-pinned product categories, lifecycle stages, and process outcomes exposed through the same diagnostic. Both packages accept multiple named instances. Loading either provider starts no product process, and each tool accepts only a standalone text task; product and instance selection remain deployment configuration.
 
 Both providers report `inheritsParentContext: false`, advertise no optional start capabilities, and pass the parent Session cwd without copying the parent conversation. Their documented tools use `backgroundMode: 'one-shot'` and `maxDepth: 'provider-managed'`: the consumer keeps foreground collection as the default and may place the same run in the generic Job runtime, while recursion policy stays with the out-of-process product. Every call creates a fresh product process and a non-resumable product conversation. `ctx.subagents` owns named-request resolution and paired lifecycle events; `dsh-tool-subagent` owns model-visible scheduling and foreground-versus-Job adaptation; `ctx.jobs` and `dsh-tool-jobs` own Job ids, state, output, controls, notices, and parent-owner cancellation; each product provider owns native result mapping, while `dsh-subprocess` owns credential scrubbing, process-tree termination, and whole-tree exit observation.
 
@@ -38,11 +38,11 @@ configured tool -> dsh-tool-subagent -> ctx.subagents -> product provider -> pro
 
 Before publication, the provider validates a non-empty text-only task, starts the managed app-server in the parent workspace, completes `initialize` → `initialized`, maps the resolved mode into official `thread/start` fields, and creates an `ephemeral: true` thread. The fixed app-server argv contains no mode or task text. The published run owns exactly one `turn/start`; its thread and turn ids remain private and are never persisted in the parent Session.
 
-`turn/completed` is the authoritative remote terminal fact. The latest `agentMessage` with `phase: "final_answer"` wins, and that selected message must contain nonblank text. When the product emits no explicit final phase, the latest message with `phase: null` is the compatibility fallback and must likewise be nonblank; commentary never replaces either answer. A failed turn with `error.codexErrorInfo: "contextWindowExceeded"` becomes `max-tokens`. A completed turn without an answer, every other failed or interrupted remote turn, malformed required fields in a recognized app-server frame, protocol closure, early process exit, or unknown server request becomes `error`; a permission-related error may additionally carry the shared safe diagnostic. This version has no native refusal terminal and therefore produces no `refusal`. Local cancellation wins its race and remains `aborted` without permission detail.
+`turn/completed` is the authoritative remote terminal fact. The latest `agentMessage` with `phase: "final_answer"` wins, and that selected message must contain nonblank text. When the product emits no explicit final phase, the latest message with `phase: null` is the compatibility fallback and must likewise be nonblank; commentary never replaces either answer. The [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns Codex error-info categories, HTTP status, lifecycle stages, process outcomes, and stop-reason preservation. Local cancellation remains `aborted` without a failure diagnostic.
 
 For command and file approvals, the unattended wire selects a non-approval decision offered by the request, preferring `cancel`; the stable 0.147.0 request shape without an offered-decision list falls back to `decline`. It grants no requested permissions for the turn, answers user-input requests with no answers, and declines MCP elicitation. It records safe categories for those requests, declined command/file items, and `sandboxError`. Codex emits some early `never` rejections and sandbox violations only on structured stderr, so the Provider pipes and forwards stderr unchanged while matching two fixed signatures in a bounded per-run tail; raw stderr never enters the diagnostic. A request with no legal unattended response, or any unknown server request, fails the run instead of waiting for a user interface the provider does not supply.
 
-An unpublished startup failure closes the wire, terminates the acquired process tree, waits for exit, detaches the stderr observer, and then rejects `start()`. Published disposal best-effort interrupts a known turn, closes the wire, ends stdin, invokes the shared termination escalation, waits for whole-tree exit, and detaches the observer. Result failure and teardown failure stay independently observable.
+An unpublished startup failure closes the wire, terminates the acquired process tree, waits for exit, detaches the stderr observer, and then rejects `start()` with its fixed operation stage. Published disposal best-effort interrupts a known turn, closes the wire, ends stdin, invokes the shared termination escalation, waits for whole-tree exit, and detaches the observer. Independent cleanup failure reports `teardown`; when startup and rollback both fail, the aggregate's top message retains both safe stage lines while the underlying causes remain internal.
 
 Codex 0.147.0 speaks the Responses protocol, while DeepSeek's public OpenAI-compatible endpoint speaks Chat Completions. The credentialed Codex e2e therefore uses a loopback-only, test-private bridge for one no-tool nonce request: real Codex sends Responses to the bridge, the bridge forwards the received bearer credential and extracted task to the fixed official DeepSeek endpoint, and it wraps the real text in the minimal Responses SSE lifecycle. The bridge is neither a production proxy nor evidence that Codex connects to DeepSeek Chat Completions natively.
 
@@ -52,9 +52,9 @@ Codex 0.147.0 speaks the Responses protocol, while DeepSeek's public OpenAI-comp
 
 The public configuration contains a non-empty `providerName`, an explicit `env` overlay, a positive finite `disposeGraceMs` no greater than the repository's shared `MAX_TIMER_DELAY_MS`, and a five-value native `permissionMode` that defaults to `dontAsk`. Each named instance retains those resolved values for its own runs. Each run creates its own `AbortController`, sets `persistSession: false`, disables `AskUserQuestion`, and passes the resolved mode to the SDK; only `bypassPermissions` receives the SDK's explicit dangerous confirmation. The provider deliberately omits `settingSources`, so the SDK reads the host's normal user, project, and local Claude settings relative to the parent Session cwd. It neither copies nor filters those settings and does not create or modify login state. Remaining permission prompts are denied, MCP elicitation is declined, and blocking dialogs fail closed instead of waiting for a user interface the provider does not own.
 
-The provider publishes only after both the SDK `Query` and a live managed CLI handle exist. It consumes the complete SDK stream and completes only when a `result` message has `subtype: "success"`, `is_error: false`, and a nonblank `result`, and the iterator then ends normally. Every SDK error subtype, an error-marked success, a missing result, iterator failure, protocol failure, or process failure becomes `error`. When a permission denial or unattended callback contributes to that failure, the result may additionally carry the bounded, non-assistant diagnostic owned by the non-interactive permissions decision. SDK turn, budget, and structured-output limits are not token-window facts, and the SDK exposes no native refusal terminal, so this provider produces neither `max-tokens` nor `refusal`. Local cancellation wins and becomes `aborted` without permission detail.
+The provider publishes only after both the SDK `Query` and a live managed CLI handle exist. It consumes the complete SDK stream and completes only when a `result` message has `subtype: "success"`, `is_error: false`, and a nonblank `result`, and the iterator then ends normally. The [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns every non-success category, stage, process outcome, and its ordering with a contributing permission decision. Local cancellation wins and becomes `aborted` without either diagnostic fact.
 
-Startup rollback and published disposal close the SDK query, abort the per-run controller, invoke shared process-tree termination, and wait for whole-tree exit. `Query.close()` expresses graceful protocol intent but does not replace the subprocess owner's exit proof. Query-close failure, process failure, and teardown failure remain independently observable.
+Startup rollback and published disposal close the SDK query, abort the per-run controller, invoke shared process-tree termination, and wait for whole-tree exit. `Query.close()` expresses graceful protocol intent but does not replace the subprocess owner's exit proof. An unpublished failure exposes only fixed `query-start` facts; a published process failure can expose its independent exit code and signal; an independent cleanup rejection exposes `teardown`. Original SDK, Host, and cleanup errors remain on internal cause chains and logs rather than entering the diagnostic.
 
 The credentialed Claude Code e2e uses the official DeepSeek Claude Code contract directly: the runtime-only DeepSeek key becomes `ANTHROPIC_AUTH_TOKEN`, the fixed official base gains `/anthropic`, and the main and subagent model variables select the documented DeepSeek models. It starts the production provider and real SDK/CLI, requires one random nonce as the complete answer, persists no credential in settings, and waits for every managed handle to exit.
 
@@ -62,11 +62,11 @@ The credentialed Claude Code e2e uses the official DeepSeek Claude Code contract
 
 Each product owns branch-complete package tests, a required keyless real-product spec, a Loader composition e2e, and a credentialed DeepSeek e2e. The keyless product tier uses the exact official distribution under test, a non-empty fake product key, an isolated temporary workspace and product home, and a loopback fixed-answer model. Missing product requests, wrong authentication, altered task text, a non-exact answer, a skipped real product, or a surviving managed handle fails the required test. The Codex Loader fixture exposes two named Codex instances and tools; the Claude Code Loader fixture exposes the default Codex tool plus two named Claude Code instances and tools. Both fixtures include generic Job controls and start neither product process. The credentialed tier starts the same production provider and real product with a runtime-only key, requires a unique nonce from the fixed official DeepSeek service, and proves quiescence again; it self-skips only when a local operator supplied no key, while trusted CI preflights the secret.
 
-The Codex evidence pins `@openai/codex@0.147.0`, `codex-cli 0.147.0`, and all six optional platform aliases. Its real-product spec observes the package-local wrapper argv, exact Bearer key, original task, byte-exact final answer, thread-level `never` overriding ambient `on-request`, automatic-review startup, unattended command rejection with safe diagnostic and no file side effect, explicit dangerous-bypass writing in suite-owned temporary storage, local cancellation, wrapper/native whole-tree exit, and missing-payload failure without host fallback.
+The Codex evidence pins `@openai/codex@0.147.0`, `codex-cli 0.147.0`, and all six optional platform aliases. Its real-product spec observes the package-local wrapper argv, exact Bearer key, original task, byte-exact final answer, native permission modes, explicit dangerous-bypass writing in suite-owned temporary storage, and wrapper/native whole-tree exit. An isolated wrapper fixture proves missing-payload failure without host fallback, two named instances retain separate environments and modes, and production never resolves a host `codex` from `PATH`. The [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns schema, failure, process-outcome, and final presentation evidence.
 
 The Codex credentialed e2e registers the production provider, starts the same real app-server, and requests one random nonce through the test-private bridge described above. It fixes the external endpoint and model, stores no credential or request payload, requires exactly one completed upstream response, compares the trimmed product answer byte-for-byte with the nonce, and waits for every managed handle to exit.
 
-The Claude Code evidence pins Agent SDK 0.3.220, Claude Code 2.1.220, and the identities and versions of all eight SDK platform packages. Its real-product spec lets the SDK select the installed payload, asserts that the shared subprocess argv begins with that package's native CLI, and observes the exact `x-api-key`, original task, byte-exact final answer, an inherited interactive host setting overridden by the safe Provider mode, denied and bypassed writes in suite-owned temporary directories, safe permission diagnostics, process failure, local cancellation, and whole-tree exit. Unit coverage proves that production never resolves host `PATH`, omits the executable override, forwards the SDK-selected Windows `claude.exe` without a batch shim, and surfaces the SDK's missing-payload error without host fallback. This evidence proves the pinned official SDK/CLI integration rather than compatibility with independently installed Claude versions. Loader coverage resolves both products through their optional Bundle patches while starting neither product.
+The Claude Code evidence pins Agent SDK 0.3.220, Claude Code 2.1.220, and all eight SDK platform packages. Its real-product spec lets the SDK select the installed payload, asserts that the shared subprocess argv begins with that package's native CLI, and observes the exact `x-api-key`, original task, byte-exact final answer, native permission modes, suite-owned denied and bypassed writes, and whole-tree exit. Package tests prove that production never resolves host `PATH`, omits the executable override, and forwards the SDK-selected Windows `claude.exe` without a batch shim. This evidence proves the pinned official SDK/CLI integration rather than compatibility with independently installed Claude versions; the [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns failure and process-outcome evidence. Loader coverage resolves both products through their optional Bundle patches while starting neither product.
 
 The Claude Code credentialed e2e maps the key and fixed official endpoint only in the provider's in-memory environment, uses the documented `deepseek-v4-pro[1m]` and `deepseek-v4-flash` model variables, and traverses the production provider, official SDK, and real CLI. It compares the trimmed result with a random nonce and proves whole-tree exit without calling the Messages API directly from the test.
 
@@ -90,6 +90,6 @@ The project owner's distribution authorization is scoped to the official `@anthr
 
 Users delegate through Profile-configured one-shot tools backed by the official product integrations. Explicit Profile installation and host-plane provider placement are owned by the [production-install exclusion decision](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md); named instance identity and tool binding are owned by the [named-instance decision](2026-08-18-product-subagent-named-instances.md); per-Preset tool exposure and foreground-default optional Job scheduling are owned by the [product one-shot background decision](2026-08-12-product-subagent-one-shot-background-tasks.md). This note's provider lifecycle keeps native settings and behavior while shared services retain the sole ownership of job settlement and process-tree quiescence.
 
-Every delegation pays for a fresh product process and independent model context. Successful product payload remains final assistant text; a failed product run may separately expose the shared safe diagnostic. Background scheduling additionally exposes generic Job ids, status, completion notices, and collection or cancellation results. Both products use Bundle-pinned platform CLIs plus native account and workspace settings and the selected Provider permission mode. Credentialed e2e runs also spend external API quota and depend on the official DeepSeek endpoint; deterministic protocol, failure, cancellation, and approval coverage remains in the keyless tier. The providers do not resume sessions, stream progress, accept new human interaction, roll back tool or file side effects, or impose a wall-clock timeout.
+Every delegation pays for a fresh product process and independent model context. Successful product payload remains final assistant text; a failed product run may separately expose the shared safe diagnostic containing provider-owned permission facts or version-pinned structured failure facts. Background scheduling additionally exposes generic Job ids, status, completion notices, and collection or cancellation results. Both products use Bundle-pinned platform CLIs plus native account and workspace settings and the selected Provider permission mode. Credentialed e2e runs also spend external API quota and depend on the official DeepSeek endpoint; deterministic protocol, failure, cancellation, and approval coverage remains in the keyless tier. The providers do not resume sessions, stream progress, accept new human interaction, roll back tool or file side effects, or impose a wall-clock timeout.
 
 Compatibility is pinned by package-level unit coverage, keyless real-product loopback tests, credentialed DeepSeek nonce tests, public Loader composition, built-package and NodeNext consumer checks, generated documentation and notices, and the repository CI matrix. A supported product or DeepSeek endpoint/model baseline change must refresh those facts; production performs no separate runtime version probe.

.agents/notes/implemented/feature/2026-08-04-claude-code-and-codex-subagent-backends.zh.md

@@ -12,7 +12,7 @@ Status: implemented
 
 ## 决策
 
-harness 交付两个同级的一次性提供方包，其默认注册名称分别为 `codex` 与 `claude-code`。本说明负责它们的产品协议、结果映射和进程生命周期；[命名实例决策](2026-08-18-product-subagent-named-instances.md)负责 Profile 选择的提供方身份与静态工具绑定，[生产安装排除决策](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md)负责各自独立的可选 Bundle 与 host plane（宿主平面）放置，[产品一次性后台任务决策](2026-08-12-product-subagent-one-shot-background-tasks.md)负责模型可见的调度选择，[非交互权限决策](2026-08-15-product-subagent-noninteractive-permissions.md)则负责各产品提供方的 Profile 模式选择与诊断生产。两个包都接受多个命名实例。加载任一提供方都不会启动产品进程，而且每个工具只接受独立文本任务；产品与实例选择仍属于部署配置。
+harness 交付两个同级的一次性提供方包，其默认注册名称分别为 `codex` 与 `claude-code`。本说明负责它们的产品协议、结果映射和进程生命周期；[命名实例决策](2026-08-18-product-subagent-named-instances.md)负责 Profile 选择的提供方身份与静态工具绑定，[生产安装排除决策](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md)负责各自独立的可选 Bundle 与 host plane（宿主平面）放置，[产品一次性后台任务决策](2026-08-12-product-subagent-one-shot-background-tasks.md)负责模型可见的调度选择，[非交互权限决策](2026-08-15-product-subagent-noninteractive-permissions.md)负责各产品提供方的 Profile 模式选择与安全权限决定，[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)则负责通过同一诊断公开锁定产品版本的类别、生命周期阶段与进程结果。两个包都接受多个命名实例。加载任一提供方都不会启动产品进程，而且每个工具只接受独立文本任务；产品与实例选择仍属于部署配置。
 
 这两个提供方都报告 `inheritsParentContext: false`，不声明任何可选的启动能力，并传递父会话 cwd，但不会复制父级对话。文档所示的工具使用 `backgroundMode: 'one-shot'` 与 `maxDepth: 'provider-managed'`：消费方默认在前台收集结果，也可把同一次运行放入通用 Job 运行时，而递归策略仍由进程外产品负责。每次调用都会创建一个全新的产品进程和一次不可续接的产品对话。`ctx.subagents` 负责具名请求解析与成对生命周期事件；`dsh-tool-subagent` 负责模型可见的调度以及前台与 Job 适配；`ctx.jobs` 和 `dsh-tool-jobs` 负责 Job id、状态、输出、控制、通知与父级 owner 取消；各产品提供方负责原生结果映射，`dsh-subprocess` 则负责凭证清洗、进程树终止以及整棵进程树的退出观测。
 
@@ -38,11 +38,11 @@ configured tool -> dsh-tool-subagent -> ctx.subagents -> product provider -> pro
 
 发布前，提供方会验证非空的纯文本任务，在父级工作区中启动受管的 app-server，完成 `initialize` → `initialized` 握手，把已解析模式映射为官方 `thread/start` 字段，并创建一个 `ephemeral: true` 线程。固定 app-server argv 不包含模式或任务文本。已发布的运行只拥有一次 `turn/start`；其线程 ID 与轮次 ID 保持私有，绝不会持久化到父会话。
 
-`turn/completed` 是权威的远端终止事实。以最后一条带有 `phase: "final_answer"` 的 `agentMessage` 为准，且选中的消息必须包含非空白文本。若产品没有发出明确的最终阶段，则以最后一条 `phase: null` 的消息作为兼容性回退，该消息也必须包含非空白文本；过程说明绝不会取代上述任一答案。带有 `error.codexErrorInfo: "contextWindowExceeded"` 的失败轮次会成为 `max-tokens`。轮次完成却没有答案、其他任何远端失败或中断轮次、已识别的 app-server 帧中必需字段格式错误、协议关闭、进程提前退出或未知的服务器请求，都会产生 `error`；权限相关错误可以额外携带共享安全诊断。本版本没有原生的拒绝终止状态，因此不会产生 `refusal`。本地取消在竞态中胜出并保持为 `aborted`，且不附带权限说明。
+`turn/completed` 是权威的远端终止事实。以最后一条带有 `phase: "final_answer"` 的 `agentMessage` 为准，且选中的消息必须包含非空白文本。若产品没有发出明确的最终阶段，则以最后一条 `phase: null` 的消息作为兼容性回退，该消息也必须包含非空白文本；过程说明绝不会取代上述任一答案。[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)负责 Codex error-info 类别、HTTP status、生命周期阶段、进程结果与终止原因保持。本地取消仍是 `aborted` 且不附带失败诊断。
 
 对于命令与文件审批，无人值守的协议连接会从请求给出的决策选项中选择一项不予批准的决策，并优先选择 `cancel`；稳定的 0.147.0 请求形态没有决策选项列表，因此回退到 `decline`。它不授予该轮次请求的任何权限，不向用户输入请求提供任何答案，并拒绝 MCP elicitation。它会记录这些请求、被拒绝的命令／文件 item 与 `sandboxError` 的安全类别。Codex 的部分早期 `never` 拒绝和 sandbox violation 只写入结构化 stderr，因此提供方会 pipe 并原样转发 stderr，同时在每次运行的有界尾部中匹配两个固定签名；原始 stderr 绝不会进入诊断。若请求在无人值守模式下没有合法响应，或是未知服务器请求，此次运行就会失败，而不会等待本提供方没有提供的用户界面。
 
-若启动在发布前失败，提供方会关闭协议连接、终止已获取的进程树、等待其退出、移除 stderr observer，然后拒绝 `start()`。对已发布的运行执行资源释放时，提供方会尽力中断已知轮次、关闭协议连接、结束标准输入、调用共享的逐级终止机制，等待整棵进程树退出，并移除 observer。结果失败与清理失败仍可彼此独立地观察。
+若启动在发布前失败，提供方会关闭协议连接、终止已获取的进程树、等待其退出、移除 stderr observer，然后用固定操作阶段拒绝 `start()`。对已发布的运行执行资源释放时，提供方会尽力中断已知轮次、关闭协议连接、结束标准输入、调用共享的逐级终止机制，等待整棵进程树退出，并移除 observer。独立清理失败会报告 `teardown`；启动与回滚同时失败时，聚合的顶层消息会保留两条安全阶段说明，而底层 cause 仍只在内部可见。
 
 Codex 0.147.0 使用 Responses 协议，而 DeepSeek 的公开 OpenAI 兼容端点使用 Chat Completions。因此，带密钥 Codex e2e 会采用一个仅限回环、仅供测试内部使用的桥接层来处理一次不使用工具的随机数请求：真实 Codex 将 Responses 发送到桥接层，桥接层把收到的 Bearer 凭据与提取出的任务转发到固定的 DeepSeek 官方端点，再将真实文本包装进最小化的 Responses SSE（Server-Sent Events）生命周期。该桥接层既不是生产代理，也不能作为 Codex 原生连接 DeepSeek Chat Completions 的证据。
 
@@ -52,9 +52,9 @@ Codex 0.147.0 使用 Responses 协议，而 DeepSeek 的公开 OpenAI 兼容端
 
 公开配置包含非空的 `providerName`、显式的 `env` 覆盖项、须为正有限值且不得大于仓库共享 `MAX_TIMER_DELAY_MS` 的 `disposeGraceMs`，以及默认使用 `dontAsk` 的五值原生 `permissionMode`。每个命名实例会为自己的运行保留这些已解析值。每次运行都会创建自己的 `AbortController`，设置 `persistSession: false`、禁用 `AskUserQuestion`，并把已解析模式传给 SDK；只有 `bypassPermissions` 会取得 SDK 的显式危险确认。提供方故意省略 `settingSources`，因此 SDK 会相对于父会话 cwd 读取宿主机常规的用户、项目和本地 Claude 设置。它既不复制也不过滤这些设置，也不会创建或修改登录状态。其余权限提示会被拒绝，MCP elicitation 会被拒绝，阻塞对话会快速失败，而不会等待本提供方不负责的用户界面。
 
-只有在 SDK `Query` 与受管的活动 CLI 句柄都已存在后，提供方才会发布运行。它会消费完整的 SDK 流；只有 `result` 消息具有 `subtype: "success"`、`is_error: false` 和非空白 `result`，且迭代器随后正常结束时，运行才会完成。所有 SDK 错误子类型、标记为错误的成功消息、结果缺失、迭代器失败、协议失败或进程失败都会成为 `error`。当权限拒绝或无人值守回调参与了该失败时，结果还可以携带由非交互权限决策负责的有界、非 assistant 诊断。SDK 的轮次、预算和结构化输出限制不表示 token 窗口耗尽，而且 SDK 没有原生的拒绝终止状态，因此本提供方不会产生 `max-tokens` 或 `refusal`。本地取消会胜出并成为 `aborted`，且不附带权限说明。
+只有在 SDK `Query` 与受管的活动 CLI 句柄都已存在后，提供方才会发布运行。它会消费完整的 SDK 流；只有 `result` 消息具有 `subtype: "success"`、`is_error: false` 和非空白 `result`，且迭代器随后正常结束时，运行才会完成。[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)负责所有非成功类别、阶段、进程结果，以及它们与参与失败的权限决定之间的顺序。本地取消会胜出并成为 `aborted`，且不附带这两类诊断事实。
 
-启动回滚和已发布运行的资源释放都会关闭 SDK query、中止该次运行的控制器、调用共享的进程树终止机制，并等待整棵进程树退出。`Query.close()` 表达优雅的协议关闭意图，但不能取代子进程责任方的退出证明。Query 关闭失败、进程失败和清理失败仍可彼此独立地观察。
+启动回滚和已发布运行的资源释放都会关闭 SDK query、中止该次运行的控制器、调用共享的进程树终止机制，并等待整棵进程树退出。`Query.close()` 表达优雅的协议关闭意图，但不能取代子进程责任方的退出证明。未发布失败只公开固定的 `query-start` 事实；已发布进程失败可以分别公开退出码与信号；独立清理拒绝则公开 `teardown`。原始 SDK、Host 与清理错误只保留在内部 cause 链和日志中，不进入诊断。
 
 带密钥 Claude Code e2e 直接使用官方 DeepSeek Claude Code 约定：仅在运行时提供的 DeepSeek 密钥会映射为 `ANTHROPIC_AUTH_TOKEN`，固定的官方基础 URL 会追加 `/anthropic`，主模型与 subagent 模型变量会选择文档所示的 DeepSeek 模型。该测试会启动生产提供方与真实 SDK 和 CLI，要求一个随机数作为完整答案，不会把任何凭据持久化到设置中，并等待所有受管句柄退出。
 
@@ -62,11 +62,11 @@ Codex 0.147.0 使用 Responses 协议，而 DeepSeek 的公开 OpenAI 兼容端
 
 每个产品都负责覆盖所有分支的包测试、一项必跑的无密钥真实产品测试、一项 Loader 组合 e2e 和一项带密钥 DeepSeek e2e。无密钥产品层级使用被测的确切官方发行版、非空的伪产品密钥、隔离的临时工作区与产品主目录，以及能返回固定答案的回环模型。产品请求缺失、身份验证错误、任务文本被改动、答案不完全一致、真实产品被跳过或受管句柄仍存活，都会使这项必跑测试失败。Codex Loader fixture 会公开两个命名 Codex 实例与工具；Claude Code Loader fixture 会公开默认 Codex 工具以及两个命名 Claude Code 实例与工具。两个 fixture 都包含通用 Job 控制工具，而且不会启动任何产品进程。带密钥层级会使用仅在运行时提供的密钥启动同一生产提供方与真实产品，要求从固定的 DeepSeek 官方服务取得唯一随机数，并再次证明完全停稳；仅当本地操作者未提供密钥时才会自行跳过，而受信任的 CI 会预检该 secret。
 
-Codex 证据会锁定 `@openai/codex@0.147.0`、`codex-cli 0.147.0` 与六个平台 alias。其真实产品测试会观测包内 wrapper argv、确切的 Bearer 密钥、原始任务、逐字节完全一致的最终回答、线程级 `never` 对环境中 `on-request` 的覆盖、自动评审启动、带安全诊断且不产生文件副作用的无人值守命令拒绝、测试拥有临时存储中的显式危险绕过写入、本地取消、wrapper／原生整棵进程树退出，以及载荷缺失时不回退宿主命令的失败。
+Codex 证据会锁定 `@openai/codex@0.147.0`、`codex-cli 0.147.0` 与六个平台 alias。其真实产品测试会观测包内 wrapper argv、确切的 Bearer 密钥、原始任务、逐字节完全一致的最终回答、原生权限模式、测试拥有临时存储中的显式危险绕过写入，以及 wrapper／原生整棵进程树退出。独立 wrapper fixture 会证明载荷缺失时不回退宿主命令，两个命名实例会保留彼此独立的环境与模式，生产环境也不会从 `PATH` 解析宿主 `codex`。[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)负责 schema、失败、进程结果与最终呈现证据。
 
 带密钥 Codex e2e 会注册生产提供方，启动同样的真实 app-server，并通过上述测试专用桥接层请求一个随机数。该测试固定外部端点与模型，不存储任何凭据或请求载荷，要求上游恰好完成一次响应，将去除首尾空白后的产品答案与该随机数逐字节比较，并等待所有受管句柄退出。
 
-Claude Code 证据会锁定 Agent SDK 0.3.220、Claude Code 2.1.220，以及八个 SDK 平台包的身份与版本。真实产品测试会让 SDK 选择已安装载荷，断言共享子进程 argv 以该包的原生 CLI 开头，并观测确切的 `x-api-key`、原始任务、逐字节完全一致的最终回答、安全提供方模式对继承的交互式宿主设置的覆盖、测试所拥有临时目录中的拒绝写入与 bypass 写入、安全权限诊断、进程失败、本地取消和整棵进程树退出。单元覆盖会证明生产运行从不解析宿主 `PATH`、省略可执行文件覆盖、直接转发 SDK 所选的 Windows `claude.exe` 而不经过 batch shim，并且在载荷缺失时原样暴露 SDK 错误且不回退宿主 CLI。这项证据证明锁定的官方 SDK/CLI 集成，而不证明与独立安装的 Claude 版本兼容。Loader 覆盖会通过各自的可选 Bundle patch 解析两个产品，且不会启动任一产品。
+Claude Code 证据会锁定 Agent SDK 0.3.220、Claude Code 2.1.220 与八个 SDK 平台包。真实产品测试会让 SDK 选择已安装载荷，断言共享子进程 argv 以该包的原生 CLI 开头，并观测确切的 `x-api-key`、原始任务、逐字节完全一致的最终回答、原生权限模式、测试拥有范围内的拒绝写入与 bypass 写入，以及整棵进程树退出。包测试还会证明生产运行从不解析宿主 `PATH`、省略可执行文件覆盖，并直接转发 SDK 所选的 Windows `claude.exe` 而不经过 batch shim。这项证据证明锁定的官方 SDK/CLI 集成，而不证明与独立安装的 Claude 版本兼容；[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)负责失败与进程结果证据。Loader 覆盖会通过各自的可选 Bundle patch 解析两个产品，且不会启动任一产品。
 
 带密钥 Claude Code e2e 仅在提供方的内存环境中映射密钥与固定的官方端点，把模型变量设为文档所示的 `deepseek-v4-pro[1m]` 与 `deepseek-v4-flash`，并实际经过生产提供方、官方 SDK 与真实 CLI。它将去除首尾空白后的结果与一个随机数比较，并证明整棵进程树退出，且测试不会直接调用 Messages API。
 
@@ -90,6 +90,6 @@ Claude Code 证据会锁定 Agent SDK 0.3.220、Claude Code 2.1.220，以及八
 
 用户通过由 Profile 配置、并由官方产品集成支持的一次性工具进行委派。显式 Profile 安装与 host plane 提供方放置由[生产安装排除决策](../simplification/2026-08-12-production-dsh-excludes-product-subagent-providers.md)负责；命名实例身份与工具绑定由[命名实例决策](2026-08-18-product-subagent-named-instances.md)负责；按 Preset 暴露工具以及默认前台且可选通用 Job 的调度方式由[产品一次性后台任务决策](2026-08-12-product-subagent-one-shot-background-tasks.md)负责。本说明规定的提供方生命周期会保留原生设置与行为，而共享服务继续独占作业结算与进程树完全停稳的责任。
 
-每次委派都要承担新建产品进程和独立模型上下文的开销。成功的产品载荷仍只有最终 assistant 文本；失败的产品运行可以另行公开共享安全诊断。后台调度还会额外公开通用 Job id、状态、完成通知以及收集或取消结果。两个产品都使用 Bundle 锁定的平台 CLI，并保留原生账户与工作区设置以及所选提供方权限模式。带密钥 e2e 运行还会消耗外部 API 配额，并依赖 DeepSeek 官方端点；对协议、失败、取消与审批的确定性覆盖仍由无密钥层级承担。提供方不会恢复会话、以流式方式传送进度、接受新的人工交互、回滚工具或文件副作用，也不会施加按实际经过时间触发的超时。
+每次委派都要承担新建产品进程和独立模型上下文的开销。成功的产品载荷仍只有最终 assistant 文本；失败的产品运行可以另行公开共享安全诊断，其中包含由提供方拥有的权限事实，或锁定版本产品提供的结构化失败事实。后台调度还会额外公开通用 Job id、状态、完成通知以及收集或取消结果。两个产品都使用 Bundle 锁定的平台 CLI，并保留原生账户与工作区设置以及所选提供方权限模式。带密钥 e2e 运行还会消耗外部 API 配额，并依赖 DeepSeek 官方端点；对协议、失败、取消与审批的确定性覆盖仍由无密钥层级承担。提供方不会恢复会话、以流式方式传送进度、接受新的人工交互、回滚工具或文件副作用，也不会施加按实际经过时间触发的超时。
 
 兼容性由包级单元测试覆盖率、无密钥真实产品回环测试、带密钥 DeepSeek 随机数测试、公开 Loader 组合、已构建包与 NodeNext 消费方检查、生成的文档与声明以及仓库 CI 矩阵共同锁定。更改受支持的产品基线或 DeepSeek 端点／模型基线时必须刷新这些事实；生产环境不会另行执行运行时版本探测。

.agents/notes/implemented/feature/2026-08-15-product-subagent-noninteractive-permissions.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/feature/2026-08-15-product-subagent-noninteractive-permissions.md
-2026-08-15-product-subagent-noninteractive-permissions.md: df1f0d9939e951f16070729615a3779f1f7c2ddc
-2026-08-15-product-subagent-noninteractive-permissions.zh.md: 982b4409e08a506dec828db15c8c4aa5fcc36883
+2026-08-15-product-subagent-noninteractive-permissions.md: 8788fba3492e08090dd038fc3e7377f6bd1e29cd
+2026-08-15-product-subagent-noninteractive-permissions.zh.md: 6f254930151abce23f04de4f354bf57ad81bba61

.agents/notes/implemented/feature/2026-08-15-product-subagent-noninteractive-permissions.md

@@ -44,9 +44,9 @@ The Provider overrides only those thread fields. `CODEX_HOME`, project configura
 
 ### Failure diagnostic
 
-`SubagentResult` carries an optional `diagnostic` for provider-authored, non-assistant failure detail. A Provider removes tool inputs, file contents, environment values, credentials, and raw protocol payloads before producing it. The shared out-of-process result boundary limits the complete text to 4096 UTF-8 bytes and marks truncation without splitting a character.
+`SubagentResult` carries an optional `diagnostic` for provider-authored, non-assistant failure detail. A Provider removes tool inputs, file contents, environment values, credentials, and raw protocol payloads before producing it. The shared out-of-process result boundary limits the complete text to 4096 UTF-8 bytes and marks truncation without splitting a character. The [structured failure-facts decision](2026-08-18-product-subagent-failure-facts.md) owns non-permission product categories, lifecycle stages, and process outcomes carried by the same field.
 
-Each product records only the effective mode, request category, unattended decision, and a fixed safe reason. Claude Code derives those facts from SDK callbacks and `permission_denied` messages. Codex derives them from app-server requests, declined items, `sandboxError`, and two fixed permission signatures in a bounded stderr tail; raw stderr is still forwarded to the Host but never copied into the diagnostic. A successful result returns only the strict final answer; local cancellation remains `aborted` without permission detail; an unpublished startup failure still rejects `start()`. When a permission fact contributes to a published run that settles as `error`, the Provider attaches the diagnostic without adding it to assistant output, structured output, or `subagent/end.lastAssistantMessage`.
+Each product's permission fact contains only the effective mode, request category, unattended decision, and a fixed safe reason. Claude Code derives those facts from SDK callbacks and `permission_denied` messages. Codex derives them from app-server requests, declined items, `sandboxError`, and two fixed permission signatures in a bounded stderr tail; raw stderr is still forwarded to the Host but never copied into the diagnostic. Both Providers place their structured failure line before the latest contributing permission fact. A successful result returns only the strict final answer; local cancellation remains `aborted` without permission detail; an unpublished startup failure still rejects `start()`. The Provider never adds either diagnostic fact to assistant output, structured output, or `subagent/end.lastAssistantMessage`.
 
 The foreground consumer presents the stop-reason headline, then the optional diagnostic, then any partial assistant output. The one-shot background adapter stores the same diagnostic beside the stop reason in the failed Job detail. Providers that omit the field retain their previous behavior.
 
@@ -63,7 +63,7 @@ The foreground consumer presents the stop-reason headline, then the optional dia
 
 ## Verification
 
-Package tests pin every allowed and rejected Config value, the exact SDK and app-server field mappings, dangerous confirmations, unattended terminal responses, diagnostic sanitization and UTF-8 bound, successful-result omission, concurrent-run isolation, foreground ordering, Job detail, stderr observer disposal, and process cleanup. The real Claude Agent SDK/CLI fixture proves its safe default, restricted denial, explicit bypass, and whole-tree quiescence. The real Codex app-server fixture proves that thread-level `never` overrides ambient `on-request`, automatic review starts, dangerous bypass writes only inside suite-owned temporary storage, fixed stderr signatures produce safe diagnostics, and the wrapper/native tree exits. Loader composition proves non-default modes can be published without starting either product, and keyless ACP snapshots record the shared diagnostic presentation while the model-facing product tool schemas contain no permission parameter.
+Package tests pin every allowed and rejected Config value, the exact SDK and app-server field mappings, dangerous confirmations, unattended terminal responses, diagnostic sanitization and UTF-8 bound, successful-result omission, concurrent-run isolation, foreground ordering, Job detail, stderr observer disposal, and process cleanup. The real Claude Agent SDK/CLI fixture proves its safe default, restricted denial, explicit bypass, and whole-tree quiescence. The real Codex app-server fixture proves that thread-level `never` overrides ambient `on-request`, automatic review starts, dangerous bypass writes only inside suite-owned temporary storage, fixed stderr signatures produce safe diagnostics, and the wrapper/native tree exits. Loader composition proves non-default modes can be published without starting either product, and the keyless ACP snapshot records each product's failure diagnostic through foreground and Job presentation while the model-facing product tool schemas contain no permission parameter.
 
 ## Alternatives considered
 
@@ -83,6 +83,6 @@ Package tests pin every allowed and rejected Config value, the exact SDK and app
 
 Profiles can select each product's native restricted, automatic, planning/edit-accepting where supported, or bypass behavior before the Provider starts, while both safe defaults never ask a person. Broader modes remain explicit deployment choices and retain their native sandbox consequences.
 
-Permission failures become visible to both foreground parents and one-shot background Jobs without turning infrastructure text into an assistant answer. That diagnostic can enter model context, Job notices, API projections, and Job UI through the ordinary consumer paths, so the Provider must sanitize and bound it before result settlement.
+Permission failures become visible to both foreground parents and one-shot background Jobs without turning infrastructure text into an assistant answer. The same field can also carry the separately owned structured failure facts. It can enter model context, Job notices, API projections, and Job UI through the ordinary consumer paths, so the Provider must sanitize and bound the complete text before result settlement.
 
 The change adds no product session persistence, human approval channel, dynamic permission operation, progress stream, retry policy, or rollback. Other Providers remain valid without producing a diagnostic or exposing a permission-mode Config.

.agents/notes/implemented/feature/2026-08-15-product-subagent-noninteractive-permissions.zh.md

@@ -44,9 +44,9 @@ Codex 默认使用 `never`，并接受 Codex 0.147.0 公开的三种原生非交
 
 ### 失败诊断
 
-`SubagentResult` 携带可选的 `diagnostic`，用于提供方产生且不属于 assistant 内容的失败说明。提供方在生成它之前会排除工具输入、文件内容、环境值、凭证与原始协议载荷。共享的进程外结果边界会把完整文本限制在 4096 个 UTF-8 字节以内，并在不切断字符的前提下标记截断。
+`SubagentResult` 携带可选的 `diagnostic`，用于提供方产生且不属于 assistant 内容的失败说明。提供方在生成它之前会排除工具输入、文件内容、环境值、凭证与原始协议载荷。共享的进程外结果边界会把完整文本限制在 4096 个 UTF-8 字节以内，并在不切断字符的前提下标记截断。[结构化失败事实决策](2026-08-18-product-subagent-failure-facts.md)负责由同一字段承载的非权限产品类别、生命周期阶段与进程结果。
 
-每个产品都只记录有效模式、请求类别、无人值守决定与固定的安全原因。Claude Code 从 SDK 回调和 `permission_denied` 消息取得这些事实。Codex 从 app-server 请求、被拒绝的 item、`sandboxError` 与每次运行有界 stderr 尾部中的两个固定权限签名取得事实；原始 stderr 仍会转发给 Host，但绝不会复制进诊断。成功结果只返回严格的最终答案；本地取消仍以 `aborted` 结算且不附带权限说明；未发布的启动失败仍会拒绝 `start()`。当一项权限事实参与了已经发布、最终以 `error` 结算的运行时，提供方会附加诊断，但不会把它写入 assistant 输出、结构化输出或 `subagent/end.lastAssistantMessage`。
+每个产品的权限事实都只包含有效模式、请求类别、无人值守决定与固定的安全原因。Claude Code 从 SDK 回调和 `permission_denied` 消息取得这些事实。Codex 从 app-server 请求、被拒绝的 item、`sandboxError` 与每次运行有界 stderr 尾部中的两个固定权限签名取得事实；原始 stderr 仍会转发给 Host，但绝不会复制进诊断。两个提供方都会把结构化失败行放在最新参与失败的权限事实之前。成功结果只返回严格的最终答案；本地取消仍以 `aborted` 结算且不附带权限说明；未发布的启动失败仍会拒绝 `start()`。提供方绝不会把任一诊断事实写入 assistant 输出、结构化输出或 `subagent/end.lastAssistantMessage`。
 
 前台消费方依次呈现终止原因标题、可选诊断和任何部分 assistant 输出。一次性后台适配器会在失败 Job 的 detail 中，把同一诊断与终止原因一起保存。没有填写该字段的提供方保持原有行为。
 
@@ -63,7 +63,7 @@ Codex 默认使用 `never`，并接受 Codex 0.147.0 公开的三种原生非交
 
 ## Verification
 
-包测试固定所有允许与拒绝的 Config 值、准确的 SDK 与 app-server 字段映射、危险确认、无人值守终态、诊断脱敏与 UTF-8 上限、成功结果不携带诊断、并发运行隔离、前台顺序、Job detail、stderr observer 释放和进程清理。真实 Claude Agent SDK/CLI fixture 证明其安全默认、受限拒绝、显式 bypass 与整棵进程树完全停稳。真实 Codex app-server fixture 证明线程级 `never` 覆盖环境中的 `on-request`、自动评审可以启动、危险绕过只在测试拥有的临时存储中写入、固定 stderr 签名产生安全诊断，而且 wrapper／native 进程树会退出。Loader 组装证明非默认模式可以在不启动任一产品的情况下发布；无密钥 ACP snapshot 则记录共享诊断呈现，同时面向模型的产品工具 schema 不包含权限参数。
+包测试固定所有允许与拒绝的 Config 值、准确的 SDK 与 app-server 字段映射、危险确认、无人值守终态、诊断脱敏与 UTF-8 上限、成功结果不携带诊断、并发运行隔离、前台顺序、Job detail、stderr observer 释放和进程清理。真实 Claude Agent SDK/CLI fixture 证明其安全默认、受限拒绝、显式 bypass 与整棵进程树完全停稳。真实 Codex app-server fixture 证明线程级 `never` 覆盖环境中的 `on-request`、自动评审可以启动、危险绕过只在测试拥有的临时存储中写入、固定 stderr 签名产生安全诊断，而且 wrapper／native 进程树会退出。Loader 组装证明非默认模式可以在不启动任一产品的情况下发布；无密钥 ACP snapshot 则记录每个产品的失败诊断如何经过前台与 Job 呈现，同时面向模型的产品工具 schema 不包含权限参数。
 
 ## Alternatives considered
 
@@ -83,6 +83,6 @@ Codex 默认使用 `never`，并接受 Codex 0.147.0 公开的三种原生非交
 
 Profile 可以在提供方启动前选择各产品原生的受限、自动、在产品支持时仅规划／编辑放行，或 bypass 行为，而两个安全默认值都绝不会询问人员。更宽松的模式仍是显式部署选择，并保留其原生沙箱后果。
 
-权限失败会同时到达前台父 agent 和一次性后台 Job，且不会把基础设施文本伪装成 assistant 回答。该诊断可以沿普通消费路径进入模型上下文、Job 通知、API 投影与 Job UI，因此提供方必须在结果结算前完成脱敏和限长。
+权限失败会同时到达前台父 agent 和一次性后台 Job，且不会把基础设施文本伪装成 assistant 回答。同一字段还可以承载由另一项决策负责的结构化失败事实。它可以沿普通消费路径进入模型上下文、Job 通知、API 投影与 Job UI，因此提供方必须在结果结算前对完整文本完成脱敏和限长。
 
 本改动不增加产品会话持久化、人工审批通道、动态权限操作、进度流、重试策略或回滚。其他提供方无需产生诊断或公开权限模式 Config，仍然保持合法。

.agents/notes/implemented/feature/2026-08-18-product-subagent-failure-facts.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/feature/2026-08-18-product-subagent-failure-facts.md
+2026-08-18-product-subagent-failure-facts.md: 50d8e918f288a6b8a9b90474499b2ed20731643f
+2026-08-18-product-subagent-failure-facts.zh.md: 7dc5a73637c90a1ca1123c86d754f95c68498fd9

.agents/notes/implemented/feature/2026-08-18-product-subagent-failure-facts.md

@@ -0,0 +1,87 @@
+# Agent Note: Product subagents expose bounded structured failure facts
+
+Status: implemented
+
+English | [中文](2026-08-18-product-subagent-failure-facts.zh.md)
+
+## Problem
+
+The [Claude Code and Codex product providers](2026-08-04-claude-code-and-codex-subagent-backends.md) receive structured product failures, but a published run historically flattened most of them to the shared `error` stop reason. Product logs retained detail that the foreground parent and a [one-shot background Job](2026-08-12-product-subagent-one-shot-background-tasks.md) could not use to distinguish a product limit, an execution failure, or an early process exit.
+
+Copying SDK error text, app-server payloads, or stderr into the result would expose task text, paths, environment values, credentials, or product internals. Adding shared error fields would also make the provider-neutral [subagent seam](2026-06-21-subagent-capability-seam.md) own product version vocabularies that change independently.
+
+## Decision
+
+Each product Provider owns the mapping from its pinned official error union, current operation, and managed process outcome to one fixed safe diagnostic line. `SubagentResult` remains unchanged: consumers receive the existing bounded `diagnostic` string and do not parse its product-private fields.
+
+### Safe diagnostic
+
+The structured line has this fixed order:
+
+```text
+Product subagent failure (product: <product>; stage: <stage>; category: <category>; HTTP status: <status>; exit code: <code>; signal: <signal>)
+```
+
+The Provider omits unavailable optional fields. Exit code and signal are independent facts and are each retained when observed. A contributing permission decision from the [non-interactive permissions decision](2026-08-15-product-subagent-noninteractive-permissions.md) follows the structured line; the latest safe permission fact remains operation-local. The shared result boundary limits the complete text to 4096 UTF-8 bytes.
+
+Successful results and local cancellation expose no failure fact. Raw product errors, stderr, tool input, paths, environment values, credentials, and protocol payloads never enter the diagnostic. Startup and cleanup rejections use the same safe line in their Error message. Original failures remain on internal cause chains; Provider Host logs and forwarded stderr remain product-local observation only.
+
+### Claude Code facts
+
+Agent SDK 0.3.220 defines four error subtypes: `error_during_execution`, `error_max_turns`, `error_max_budget_usd`, and `error_max_structured_output_retries`. The Claude Code Provider preserves each exact subtype as the category while keeping the shared stop reason `error`. An error-marked or blank success uses `invalid-success`, a missing result uses `missing-result`, a process exit before an SDK terminal result uses `process-exit`, and an unrecognized value or exception uses `unknown` without copying the value.
+
+| Stage | Owned operation | Observable failure |
+| --- | --- | --- |
+| `query-start` | SDK query construction, native platform-payload startup, and unpublished rollback | `start()` rejects with fixed safe facts and any process outcome observed before rollback |
+| `query-run` | Published SDK message iteration and strict terminal-result validation | The run resolves as `error` with the exact known subtype or a fixed result category |
+| `process` | Managed CLI exits before the SDK supplies a terminal result | The run resolves as `error` with `process-exit` and the available exit code and signal |
+| `teardown` | Query close and managed process-tree release | `dispose()` rejects independently with fixed safe facts after cleanup still reaches its final exit wait |
+
+### Codex facts
+
+Codex app-server 0.147.0 defines eleven string categories and five object variants. The Provider preserves `contextWindowExceeded`, `sessionBudgetExceeded`, `usageLimitExceeded`, `serverOverloaded`, `cyberPolicy`, `internalServerError`, `unauthorized`, `badRequest`, `threadRollbackFailed`, `sandboxError`, and `other`. It also preserves `httpConnectionFailed`, `responseStreamConnectionFailed`, `responseStreamDisconnected`, `responseTooManyFailedAttempts`, and `activeTurnNotSteerable`; the four connection/stream variants retain numeric `httpStatusCode`, while the active-turn variant does not expose `turnKind`. Unknown strings, objects with another variant set, malformed values, and unclassified exceptions use `unknown`.
+
+| Stage | Owned operation | Observable failure |
+| --- | --- | --- |
+| `initialize` | App-server spawn and initialize/initialized handshake | `start()` rejects with fixed safe facts and any process outcome already observed |
+| `thread-start` | Ephemeral `thread/start` request and response validation | `start()` rejects with the thread stage and any available process outcome |
+| `turn-start` | Published `turn/start` request, provisional ids, and early frames | The run resolves as `error` with a safe unknown fallback when no structured category exists |
+| `turn` | Terminal notification, final-answer selection, and error-info mapping | The complete category and optional HTTP status reach the non-completed result |
+| `process` | Managed app-server exits before another terminal path settles | The run resolves as `error` with `process-exit` and any available code and signal |
+| `teardown` | Wire close and process-tree release | `dispose()` rejects independently; startup rollback aggregation exposes both startup and teardown lines |
+
+`contextWindowExceeded` remains `max-tokens`; every other known or unknown Codex category remains `error`, and `cyberPolicy` does not become `refusal`.
+
+### Ownership and lifecycle
+
+| Fact or resource | Owner | Consumer behavior |
+| --- | --- | --- |
+| Product error category | Pinned official SDK or app-server version | The Provider maps only the declared structured union and uses `unknown` outside it |
+| Current failure stage | Product Provider operation | Derived at the failure site; never persisted or used as a recovery state |
+| Exit code and signal | `dsh-subprocess` process handle | The Provider displays observed values without inferring missing ones |
+| Diagnostic bytes and delivery | `dsh-subagent`, foreground tool, and Job runtime | The same bounded text is presented separately from assistant output in both scheduling modes |
+| Raw product failure | Product runtime, internal cause chain, and Host observation | It remains internal and never becomes model-visible result text |
+
+## Verification
+
+Claude Code package tests pin all four SDK subtypes, invalid success, missing result, unknown values and exceptions, all four stages, independent exit code and signal fields, permission-fact ordering, sanitization, successful-result and cancellation omission, concurrent-run isolation, and cleanup completion. Codex package tests pin all sixteen error-info variants, HTTP status presence and absence, all six stages, unknown fallback, stop-reason preservation, permission ordering, sanitization, cancellation, concurrency, and cleanup aggregation. The real SDK/CLI fixture produces an actual Claude `error_max_turns`; the real app-server fixture produces an actual Codex `internalServerError`; both fixtures cover process/protocol failure and whole-tree quiescence. The keyless ACP snapshot records each product's exact diagnostic in foreground error output, a background completion notice, and `job_output`.
+
+## Alternatives considered
+
+**Return raw SDK errors, app-server payloads, or stderr.** These values can contain commands, paths, workspace content, environment values, credentials, or upstream prose. A fixed allowlisted mapping preserves actionable facts without expanding the model-visible trust boundary.
+
+**Add a shared product-error enum or structured result fields.** Claude Code and Codex version their error unions independently. A shared enum would duplicate those authorities and force unrelated Providers and consumers to track product releases.
+
+**Parse generic stderr and exception messages.** Free-form text is neither stable nor safe. Only pinned structured product fields and the managed process outcome qualify as diagnostic input.
+
+**Persist stages or add a recovery controller.** The stage is derived from the current call site only when a failure is reported. Persistence, retries, resume, and remediation need separate ownership and user contracts.
+
+**Map product limits to new shared stop reasons.** Claude Code turn and budget limits are not token-window exhaustion, and an error category does not establish refusal semantics. Existing stop reasons remain unchanged.
+
+## Consequences
+
+The parent can distinguish important Claude Code limits and Codex budget, usage, service, policy, request, connection, stream, rollback, sandbox, and active-turn failures without receiving raw product text. Foreground and background scheduling preserve the same fact because both consume one `SubagentResult`.
+
+The diagnostic is display text rather than a new public protocol. Callers may present it but must not branch on its punctuation or product-private category names. A pinned product-version upgrade must update the Provider mapping and evidence when its official error union changes.
+
+This decision adds no product session persistence, retry policy, recovery state, stderr classifier, authentication or configuration taxonomy, progress stream, or human interaction path.

.agents/notes/implemented/feature/2026-08-18-product-subagent-failure-facts.zh.md

@@ -0,0 +1,87 @@
+# Agent Note: 产品 subagent 公开有界结构化失败事实
+
+Status: implemented
+
+[English](2026-08-18-product-subagent-failure-facts.md) | 中文
+
+## Problem
+
+[Claude Code 与 Codex 产品提供方](2026-08-04-claude-code-and-codex-subagent-backends.md)会收到结构化产品失败，但已发布运行以往会把其中大多数压成共享的 `error` 终止原因。产品日志保留了细节，前台父 agent 与[一次性后台 Job](2026-08-12-product-subagent-one-shot-background-tasks.md)却无法据此区分产品限制、执行失败或进程提前退出。
+
+若把 SDK 错误文本、app-server payload 或 stderr 复制进结果，就会暴露任务文本、路径、环境值、凭证或产品内部信息。若增加共享错误字段，又会让提供方无关的 [subagent seam](2026-06-21-subagent-capability-seam.md)拥有彼此独立变化的产品版本词汇。
+
+## Decision
+
+每个产品提供方分别拥有从锁定版本官方错误联合、当前操作和受管进程结果到一行固定安全诊断的映射。`SubagentResult` 保持不变：消费方仍接收现有的有界 `diagnostic` 字符串，而且不解析其中由产品私有的字段。
+
+### 安全诊断
+
+结构化行采用以下固定顺序：
+
+```text
+Product subagent failure (product: <product>; stage: <stage>; category: <category>; HTTP status: <status>; exit code: <code>; signal: <signal>)
+```
+
+提供方会省略不可用的可选字段。退出码与信号是相互独立的事实，只要已观测到就分别保留。来自[非交互权限决策](2026-08-15-product-subagent-noninteractive-permissions.md)且参与失败的权限决定会跟在结构化行之后；最新的安全权限事实仍只属于当前操作。共享结果边界会把完整文本限制在 4096 个 UTF-8 字节以内。
+
+成功结果与本地取消都不公开失败事实。原始产品错误、stderr、工具输入、路径、环境值、凭证和协议 payload 绝不会进入诊断。启动与清理拒绝会在 Error 消息中使用同一安全行。原始失败保留在内部 cause 链中；提供方 Host 日志与转发的 stderr 也只作为产品本地观测。
+
+### Claude Code 事实
+
+Agent SDK 0.3.220 定义四种错误子类型：`error_during_execution`、`error_max_turns`、`error_max_budget_usd` 和 `error_max_structured_output_retries`。Claude Code 提供方会把每种准确子类型保留为类别，同时维持共享终止原因 `error`。标记为错误或内容空白的成功消息使用 `invalid-success`，缺失结果使用 `missing-result`，SDK 给出终态结果前发生的进程退出使用 `process-exit`，无法识别的值或异常使用 `unknown`，且不会复制原值。
+
+| 阶段 | 归属操作 | 可观察失败 |
+| --- | --- | --- |
+| `query-start` | SDK query 构造、原生平台载荷启动与未发布回滚 | `start()` 以固定安全事实和回滚前已观测到的进程结果拒绝 |
+| `query-run` | 已发布 SDK 消息迭代与严格终态结果校验 | 运行以 `error` 兑现，并携带准确已知子类型或固定结果类别 |
+| `process` | SDK 提供终态结果之前受管 CLI 已退出 | 运行以 `error` 兑现，并携带 `process-exit` 以及可用的退出码和信号 |
+| `teardown` | Query 关闭与受管进程树释放 | `dispose()` 独立拒绝并携带固定安全事实，同时清理仍会完成最终退出等待 |
+
+### Codex 事实
+
+Codex app-server 0.147.0 定义十一种字符串类别与五种对象 variant。提供方会保留 `contextWindowExceeded`、`sessionBudgetExceeded`、`usageLimitExceeded`、`serverOverloaded`、`cyberPolicy`、`internalServerError`、`unauthorized`、`badRequest`、`threadRollbackFailed`、`sandboxError` 和 `other`。它还会保留 `httpConnectionFailed`、`responseStreamConnectionFailed`、`responseStreamDisconnected`、`responseTooManyFailedAttempts` 与 `activeTurnNotSteerable`；四种连接／stream variant 会保留数值 `httpStatusCode`，而 active-turn variant 不公开 `turnKind`。未知字符串、同时含其他 variant 的对象、格式错误值与未分类异常统一使用 `unknown`。
+
+| 阶段 | 归属操作 | 可观察失败 |
+| --- | --- | --- |
+| `initialize` | App-server spawn 与 initialize/initialized 握手 | `start()` 以固定安全事实和已经观测到的进程结果拒绝 |
+| `thread-start` | 临时 `thread/start` 请求与响应校验 | `start()` 以线程阶段和可用进程结果拒绝 |
+| `turn-start` | 已发布 `turn/start` 请求、暂定 id 与早到 frame | 没有结构化类别时，运行以 `error` 和安全 unknown 回退兑现 |
+| `turn` | 终态通知、最终答案选择与 error-info 映射 | 完整类别与可选 HTTP status 进入非完成结果 |
+| `process` | 受管 app-server 在另一终态路径结算前退出 | 运行以 `error` 兑现，并携带 `process-exit` 以及可用的退出码与信号 |
+| `teardown` | Wire 关闭与进程树释放 | `dispose()` 独立拒绝；启动回滚聚合会同时公开启动与 teardown 两行 |
+
+`contextWindowExceeded` 仍是 `max-tokens`；其他所有已知或未知 Codex 类别仍是 `error`，`cyberPolicy` 不会变成 `refusal`。
+
+### 所有权与生命周期
+
+| 事实或资源 | Owner | 消费方行为 |
+| --- | --- | --- |
+| 产品错误类别 | 锁定版本的官方 SDK 或 app-server | 提供方只映射已声明的结构化联合，并对联合外值使用 `unknown` |
+| 当前失败阶段 | 产品提供方操作 | 只在失败点派生；绝不持久化，也不作为恢复状态 |
+| 退出码与信号 | `dsh-subprocess` 进程句柄 | 提供方展示已观测值，不推测缺失值 |
+| 诊断字节与送达 | `dsh-subagent`、前台工具与 Job 运行时 | 两种调度模式都把同一份有界文本与 assistant 输出分开呈现 |
+| 原始产品失败 | 产品运行时、内部 cause 链与 Host 观测 | 只保留在内部，绝不成为模型可见的结果文本 |
+
+## Verification
+
+Claude Code 包测试固定四种 SDK 子类型、无效成功、缺失结果、未知值与异常、四个阶段、相互独立的退出码与信号字段、权限事实顺序、脱敏、成功结果与取消时省略诊断、并发运行隔离和清理完成。Codex 包测试固定全部十六种 error-info variant、HTTP status 存在与缺失、六个阶段、unknown 回退、终止原因保持不变、权限顺序、脱敏、取消、并发与清理聚合。真实 SDK/CLI fixture 会产生真实的 Claude `error_max_turns`，真实 app-server fixture 会产生真实的 Codex `internalServerError`；两个 fixture 都覆盖进程／协议失败与整棵进程树完全停稳。无密钥 ACP snapshot 会在前台错误输出、后台完成通知和 `job_output` 中记录两个产品各自的准确诊断。
+
+## Alternatives considered
+
+**返回原始 SDK 错误、app-server payload 或 stderr。** 这些值可能包含命令、路径、工作区内容、环境值、凭证或上游文本。固定白名单映射可以保留可操作事实，同时不扩大模型可见的信任边界。
+
+**增加共享产品错误 enum 或结构化结果字段。** Claude Code 与 Codex 各自独立版本化错误联合。共享 enum 会复制这些权威，并迫使无关提供方和消费方跟随产品版本。
+
+**解析通用 stderr 与异常消息。** 自由文本既不稳定也不安全。只有锁定版本产品提供的结构化字段和受管进程结果可以成为诊断输入。
+
+**持久化阶段或增加恢复控制器。** 阶段只在报告失败时从当前调用点派生。持久化、重试、resume 与修复需要独立的所有权和用户约定。
+
+**把产品限制映射为新的共享终止原因。** Claude Code 的轮次和预算限制并不表示 token 窗口耗尽，错误类别也不能证明拒绝语义。既有终止原因保持不变。
+
+## Consequences
+
+父 agent 可以区分重要的 Claude Code 限制，以及 Codex 预算、用量、服务、策略、请求、连接、stream、回滚、sandbox 和 active-turn 失败，而不会收到原始产品文本。前台与后台调度会保留同一事实，因为二者都消费同一个 `SubagentResult`。
+
+诊断只是展示文本，不是新的公开协议。调用方可以呈现它，但不得根据其标点或产品私有类别名称进行分支。锁定产品版本升级并改变官方错误联合时，必须同步更新提供方映射与证据。
+
+本决策不增加产品会话持久化、重试策略、恢复状态、stderr 分类器、身份验证或配置分类体系、进度流或人工交互路径。

.agents/notes/implemented/feature/2026-08-18-web-home-path-tilde.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/feature/2026-08-18-web-home-path-tilde.md
+2026-08-18-web-home-path-tilde.md: b148833bab09eadce4c9c1a362dd99d04eba5977
+2026-08-18-web-home-path-tilde.zh.md: 9d15cd6dca1128927389d5731dff6bf831cffe76

.agents/notes/implemented/feature/2026-08-18-web-home-path-tilde.md

@@ -0,0 +1,37 @@
+# Agent Note: Web UI abbreviates POSIX home paths as `~`
+
+Status: implemented
+
+English | [中文](2026-08-18-web-home-path-tilde.zh.md)
+
+## Problem
+
+Workspace hover cards and Tool call summaries showed full POSIX home paths. Those strings are long, repeat the same prefix on every row, and make the sidebar and transcript harder to scan. Windows paths must stay verbatim because `~` is not a Windows filesystem convention.
+
+## Decision
+
+`host.describe` reports the host account `home` as a required field. Client and Host ship together, so the field is required rather than optional. ApiProxy fills it from `homedir()` at describe time.
+
+`abbreviateHomePath` in `dsh-client-runtime` is the display-only helper. It returns `~` or `~/…` when the path is the POSIX home or a descendant, and leaves the path unchanged when `home` is missing, empty, or `/`, when either value is a Windows drive or UNC path, or when the match is only a prefix (`/Users/u` does not claim `/Users/u2`). Tool summaries run workspace-relative shortening first, then this helper, so a path inside the session cwd stays short. `filePath`, Host open, and Workspace hover copy keep the authored filesystem path.
+
+`ui-tool` and `ui-workspace` inject `connection.hostDescription` at their own slot registrations. ChatView does not grow a Host-description hook. The field is required on `ConnectionHandle`; test fakes supply a source whose snapshot may be undefined before connect.
+
+The fixture Host home is `/home/fixture`. A second fixture Workspace at `/home/fixture/Documents/project` lets assembled replay hover `~/Documents/project` without moving the existing `/tmp/fixture` account. TerminalBlock's own prompt-label collapse is unchanged.
+
+## Alternatives considered
+
+**Guess `/Users` or `/home` without the real home.** Rejected because a shared prefix is not an account home, and `/Users/shared` or `/home/src` would abbreviate incorrectly.
+
+**Abbreviate Windows `%USERPROFILE%` as `~` as well.** Rejected because the acceptance rule keeps Windows paths verbatim, and `~` is not how Explorer or `cmd` spell those paths.
+
+**Put the helper in `dsh-home-paths`.** Rejected because that package expands configuration tildes on Node; this helper is a browser display rewrite and must not pull Node `os` into client bundles.
+
+**Thread `home` from ChatView owner props.** Rejected because it enlarges the conversation inject face and every ChatView test harness for a display fact only Tool and Workspace cards consume.
+
+## Consequences
+
+POSIX home-rooted Workspace hover paths and leftover Tool path summaries display as `~`. Copy and open still use the full path. Windows drive and UNC paths never become `~`. A Host that reports `/` as home does not turn the whole filesystem into `~`. Before the first describe, or while reconnecting, the source snapshot is undefined and paths stay unabbreviated.
+
+## Testing
+
+Package tests cover `abbreviateHomePath`, `toolRowModel` / `readCardModel` home abbreviation, Workspace hover display versus copy, and `host.describe` schema plus live `homedir()`. Assembled replay `apps/web/tests/home-path-tilde.snapshot.ts` hovers the fixture home-descendant Workspace. Product-GUI PRs still record a real-browser GIF of the hover card.

.agents/notes/implemented/feature/2026-08-18-web-home-path-tilde.zh.md

@@ -0,0 +1,37 @@
+# Agent Note: Web UI abbreviates POSIX home paths as `~`
+
+Status: implemented
+
+[English](2026-08-18-web-home-path-tilde.md) | 中文
+
+## Problem
+
+Workspace 悬停卡片和 Tool 调用摘要会显示完整的 POSIX 家目录路径。这些字符串很长，每行重复同一前缀，侧边栏和对话记录更难扫读。Windows 路径必须保持原样，因为 `~` 不是 Windows 文件系统约定。
+
+## Decision
+
+`host.describe` 把宿主账户的 `home` 作为必填字段上报。Client 与 Host 一同发布，因此该字段是必填而不是可选。ApiProxy 在 describe 时用 `homedir()` 填入。
+
+`dsh-client-runtime` 中的 `abbreviateHomePath` 是仅用于展示的辅助函数。当路径是 POSIX 家目录或其后代时返回 `~` 或 `~/…`；`home` 缺失、为空或为 `/`，任一侧是 Windows 盘符或 UNC 路径，或只是前缀命中（`/Users/u` 不能收走 `/Users/u2`）时，路径保持不变。Tool 摘要先做工作区相对缩短，再调用该辅助函数，因此会话 cwd 内的路径仍然更短。`filePath`、Host 打开以及 Workspace 悬停复制仍使用作者给出的文件系统路径。
+
+`ui-tool` 与 `ui-workspace` 在各自的 slot 注册上注入 `connection.hostDescription`。ChatView 不增加 Host 描述钩子。该字段在 `ConnectionHandle` 上是必填的；测试假对象提供一个来源，其快照在连接完成前可以为 undefined。
+
+fixture 的 Host 家目录是 `/home/fixture`。第二个 fixture Workspace 位于 `/home/fixture/Documents/project`，组装回放可以悬停出 `~/Documents/project`，而不必移动现有的 `/tmp/fixture` 账户。TerminalBlock 自有的提示符标签折叠保持不变。
+
+## Alternatives considered
+
+**在没有真实 home 的情况下猜测 `/Users` 或 `/home`。** 否决，因为共享前缀不是账户家目录，`/Users/shared` 或 `/home/src` 会被错误缩写。
+
+**同样把 Windows `%USERPROFILE%` 缩写成 `~`。** 否决，因为验收规则要求 Windows 路径保持原样，而且 Explorer 与 `cmd` 并不这样拼写这些路径。
+
+**把辅助函数放进 `dsh-home-paths`。** 否决，因为该包在 Node 上展开配置里的波浪号；本辅助函数是浏览器展示改写，不能把 Node `os` 拉进 client 包。
+
+**从 ChatView owner props 向下传递 `home`。** 否决，因为它会扩大 conversation 注入面和每一份 ChatView 测试夹具，而只有 Tool 与 Workspace 卡片消费这个展示事实。
+
+## Consequences
+
+POSIX 家目录下的 Workspace 悬停路径，以及缩短 cwd 后仍落在家目录里的 Tool 路径摘要，会显示为 `~`。复制与打开仍使用完整路径。Windows 盘符和 UNC 路径永远不会变成 `~`。若 Host 把 `/` 报成 home，不会把整个文件系统收成 `~`。首次 describe 之前或重连期间，来源快照为 undefined，路径保持未缩写。
+
+## Testing
+
+包测试覆盖 `abbreviateHomePath`、`toolRowModel`／`readCardModel` 的家目录缩写、Workspace 悬停展示与复制，以及 `host.describe` schema 与实时 `homedir()`。组装回放 `apps/web/tests/home-path-tilde.snapshot.ts` 悬停 fixture 中位于家目录下的 Workspace。面向产品 GUI 的 PR 仍需录制悬停卡片的真实浏览器 GIF。

.github/workflows/build-exe-for-python-sdk.yml

@@ -261,17 +261,20 @@ jobs:
           name: deepseek_harness_sdk-${{ needs.plan.outputs.version }}-py3-none-any.whl
           path: dist-python
 
-      - name: Install only the SDK into a clean venv and run zero-config
+      - name: Install local SDK and runtime wheels into a clean venv
         env:
-          VERSION: ${{ needs.plan.outputs.version }}
+          RUNTIME_WHEEL: ${{ steps.runtime.outputs.wheel }}
+          SDK_WHEEL: deepseek_harness_sdk-${{ needs.plan.outputs.version }}-py3-none-any.whl
         run: |
           set -euo pipefail
           python -m venv "$RUNNER_TEMP/dsh-sdk-smoke"
           "$RUNNER_TEMP/dsh-sdk-smoke/bin/python" -m pip install \
-            --find-links dist-python \
-            deepseek-harness-sdk=="$VERSION"
+            "dist-python/$SDK_WHEEL" \
+            "dist-python/$RUNTIME_WHEEL"
           "$RUNNER_TEMP/dsh-sdk-smoke/bin/python" scripts/smoke-python-runtime.py \
             --scenario sdk-default
+          "$RUNNER_TEMP/dsh-sdk-smoke/bin/python" scripts/smoke-python-runtime.py \
+            --scenario sdk-mcp
 
       - name: Check Linux GLIBC requirements
         if: runner.os == 'Linux'
@@ -297,7 +300,8 @@ jobs:
         if: runner.os == 'Linux'
         env:
           RUNNER_ARCH: ${{ runner.arch }}
-          VERSION: ${{ needs.plan.outputs.version }}
+          RUNTIME_WHEEL: ${{ steps.runtime.outputs.wheel }}
+          SDK_WHEEL: deepseek_harness_sdk-${{ needs.plan.outputs.version }}-py3-none-any.whl
         run: |
           set -euo pipefail
           case "$RUNNER_ARCH" in
@@ -305,10 +309,11 @@ jobs:
             ARM64) image=quay.io/pypa/manylinux_2_28_aarch64 ;;
             *) echo "::error::Unsupported Linux runner architecture $RUNNER_ARCH"; exit 1 ;;
           esac
-          docker run --rm -e VERSION -e DSH_TELEMETRY_DISABLED -v "$PWD:/work" -w /work "$image" bash -euxo pipefail -c '
+          docker run --rm -e RUNTIME_WHEEL -e SDK_WHEEL -e DSH_TELEMETRY_DISABLED -v "$PWD:/work" -w /work "$image" bash -euxo pipefail -c '
             /opt/python/cp310-cp310/bin/python -m venv /tmp/dsh-sdk
-            /tmp/dsh-sdk/bin/python -m pip install --find-links /work/dist-python deepseek-harness-sdk=="$VERSION"
+            /tmp/dsh-sdk/bin/python -m pip install "/work/dist-python/$SDK_WHEEL" "/work/dist-python/$RUNTIME_WHEEL"
             /tmp/dsh-sdk/bin/python /work/scripts/smoke-python-runtime.py --scenario sdk-default
+            /tmp/dsh-sdk/bin/python /work/scripts/smoke-python-runtime.py --scenario sdk-mcp
           '
 
       - uses: actions/upload-artifact@v7

.github/workflows/python-release.yml

@@ -68,10 +68,15 @@ jobs:
           print(f"version={release['pep440_version'](repository_version)}")
           PY
 
-      - name: Install and run the published entry path
+      - name: Install local release wheels and run the public entry path
+        env:
+          VERSION: ${{ steps.compatibility-version.outputs.version }}
         run: |
-          python -m pip install --find-links dist "deepseek-harness-sdk==${{ steps.compatibility-version.outputs.version }}"
+          python -m pip install \
+            "dist/deepseek_harness_sdk-$VERSION-py3-none-any.whl" \
+            "dist/deepseek_harness_runtime_bin-$VERSION-py3-none-manylinux_2_28_x86_64.whl"
           python scripts/smoke-python-runtime.py --scenario sdk-default
+          python scripts/smoke-python-runtime.py --scenario sdk-mcp
 
   validate:
     name: Validate release candidate

.gitlab-ci.yml

@@ -45,6 +45,7 @@ sdk-wheel:
     - python -m venv .wheel-smoke
     - .wheel-smoke/bin/python -m pip install --find-links "release/$PLATFORM" --find-links release/sdk deepseek-harness-sdk=="$DSH_WHEEL_VERSION"
     - .wheel-smoke/bin/python scripts/smoke-python-runtime.py --scenario sdk-default
+    - .wheel-smoke/bin/python scripts/smoke-python-runtime.py --scenario sdk-mcp
     - |
       if [ "${PLATFORM#linux-}" != "$PLATFORM" ]; then
         readelf --version-info "$EXE" > glibc-versions.txt
@@ -56,7 +57,7 @@ sdk-wheel:
           linux-arm64) image=quay.io/pypa/manylinux_2_28_aarch64 ;;
           *) echo "Unsupported Linux platform $PLATFORM"; exit 1 ;;
         esac
-        docker run --rm -v "$PWD:/work" -w /work "$image" bash -euxo pipefail -c "/opt/python/cp310-cp310/bin/python -m venv /tmp/dsh-sdk && /tmp/dsh-sdk/bin/python -m pip install --find-links /work/release/$PLATFORM --find-links /work/release/sdk deepseek-harness-sdk==$DSH_WHEEL_VERSION && /tmp/dsh-sdk/bin/python /work/scripts/smoke-python-runtime.py --scenario sdk-default"
+        docker run --rm -v "$PWD:/work" -w /work "$image" bash -euxo pipefail -c "/opt/python/cp310-cp310/bin/python -m venv /tmp/dsh-sdk && /tmp/dsh-sdk/bin/python -m pip install --find-links /work/release/$PLATFORM --find-links /work/release/sdk deepseek-harness-sdk==$DSH_WHEEL_VERSION && /tmp/dsh-sdk/bin/python /work/scripts/smoke-python-runtime.py --scenario sdk-default && /tmp/dsh-sdk/bin/python /work/scripts/smoke-python-runtime.py --scenario sdk-mcp"
       fi
     - |
       if [ "$PLATFORM" = macos-arm64 ]; then

apps/web/tests/home-path-tilde.snapshot.ts

@@ -0,0 +1,45 @@
+// @vitest-environment jsdom
+// Assembled POSIX home-path display: the fixture Host home is `/home/fixture`
+// and a second Workspace lives under it. The sidebar hover card must show
+// `~/Documents/project` while copy still writes the full path.
+import { mkdirSync, writeFileSync } from 'node:fs'
+import { dirname, join } from 'node:path'
+import { act, fireEvent, screen, waitFor, within } from '@testing-library/react'
+import { describe, expect, it } from 'vitest'
+import { installAssembledBootEnv, mountAssembledApp, REFRESHING_GOLDEN } from './assembled-boot.ts'
+
+const EXPECTED = join(process.cwd(), 'apps/web/tests/snapshots/home-path-tilde/workspace-hover.expected.txt')
+
+installAssembledBootEnv()
+
+describe('assembled POSIX home-path display', () => {
+  it('shows the home-descendant Workspace path as ~ and copies the full path', async () => {
+    mountAssembledApp()
+
+    const tree = await screen.findByRole('tree', { name: 'Sessions' }, { timeout: 10_000 })
+    const group = (await within(tree).findAllByText('project'))
+      .map(el => el.closest<HTMLElement>('[role="treeitem"]'))
+      .find(el => el?.getAttribute('aria-expanded') !== null)
+    if (group == null) throw new Error('home-descendant Workspace group missing')
+
+    fireEvent.pointerEnter(group.parentElement as HTMLElement)
+    const hoverPath = await waitFor(() => {
+      const found = screen.getByText('~/Documents/project')
+      expect(found).toBeTruthy()
+      return found
+    }, { timeout: 2_000 })
+    expect(screen.queryByText('/home/fixture/Documents/project')).toBeNull()
+    const copy = screen.getByRole('button', { name: 'Copy: /home/fixture/Documents/project' })
+
+    const shape = [
+      `hover=${hoverPath.textContent}`,
+      `copy=${copy.getAttribute('aria-label')}`,
+    ].join('\n') + '\n'
+    if (REFRESHING_GOLDEN) {
+      mkdirSync(dirname(EXPECTED), { recursive: true })
+      writeFileSync(EXPECTED, shape)
+    }
+    await expect(shape).toMatchFileSnapshot(EXPECTED)
+    act(() => { fireEvent.pointerLeave(group.parentElement as HTMLElement) })
+  })
+})

apps/web/tests/message-feedback-layout.e2e.ts

@@ -0,0 +1,336 @@
+// Web e2e scenario: with the feedback note editor open, the assistant IconActions
+// row stays one intact line (no wrapping, nothing pushed out), and the note
+// editor floats above the transcript in a popover that escapes the conversation
+// column's overflow clip and stays inside the viewport.
+//
+// The hazard this pins: a slot-contributed note editor (260px textarea plus
+// Save and Cancel) cannot fit the shared IconActions row at ANY viewport, and an
+// inline expansion made the row wider than the column — full-screen desktop
+// included — so the branch action and the clock were pushed out of view by later
+// flex items. The fix is to not mount the editor in the row at all: it is a
+// popover portaled to document.body and fixed-positioned from the note trigger's
+// rect, so the row keeps its single 28px line of icons and the trigger, and the
+// panel cannot be cropped by the column's overflow because it lives outside it.
+//
+// The sweep records, per viewport, whether the open editor keeps the actions row
+// on one line with zero overflow, whether the panel is outside the column (proof
+// it escapes the clip), whether the panel stays inside the viewport (proof the
+// clamp works), and whether it sits by its trigger. All relations, no absolute
+// pixels: the column width follows the viewport, the sidebar, and the platform's
+// scrollbar, so a golden carrying pixels would document the platform, not the
+// behavior.
+//
+// Zero model calls: a settled transcript is cold-seeded, so nothing streams.
+import { readFile } from 'node:fs/promises'
+import { fileURLToPath } from 'node:url'
+import { join } from 'node:path'
+import type { Browser, Page } from 'playwright'
+import { chromium } from 'playwright'
+import { afterAll, beforeAll, describe, expect, it, onTestFailed } from 'vitest'
+import {
+  compareOrRefreshGolden, launchWebScaffold, seedSession, watchConsole, webSnapshotMode,
+  type WebScaffold,
+} from './scaffold.ts'
+import { newEnglishPage, saveFailureShot } from './support.ts'
+
+const SNAPSHOT_DIR = fileURLToPath(new URL('./snapshots/message-feedback-layout', import.meta.url))
+/**
+ * Committed golden of the popover relations at every stop. Booleans and counts
+ * only, never absolute coordinates.
+ */
+const GEOMETRY_EXPECTED = join(SNAPSHOT_DIR, 'geometry.expected.md')
+const MODE = webSnapshotMode()
+/** Borrowed read-only: this scenario needs any settled assistant message to rate. */
+const SEED = fileURLToPath(new URL('./snapshots/seeded-history/seed.jsonl', import.meta.url))
+const SEED_ID = 'message-feedback-layout-e2e'
+/** Viewport widths from full-screen desktop down to a narrow window. */
+const WIDTHS = [1680, 1280, 1024, 900, 700, 600]
+
+/** One viewport stop: how the row reads with the note editor closed and open, plus the popover's own relations. */
+export interface PopoverMetrics {
+  /** Viewport width the stop was measured at. */
+  width: number
+  /** The row's scrollable overflow with the note editor closed (natural row width). */
+  rowOverflowClosed: number
+  /** The row's scrollable overflow with the note editor open; must equal the closed value. */
+  rowOverflowOpen: number
+  /** Flex lines the row occupies with the note editor open; the editor must not reflow it. */
+  rowLines: number
+  /** Row items whose right edge escapes the column, editor closed. */
+  itemsOutsideColumnClosed: number
+  /** Row items whose right edge escapes the column, editor open; must equal the closed value. */
+  itemsOutsideColumnOpen: number
+  /** True when the portaled panel is NOT inside the column (escapes its overflow clip). */
+  panelOutsideColumn: boolean
+  /** True when the panel lies fully inside the viewport (the clamp holds). */
+  panelWithinViewport: boolean
+  /** Horizontal separation between the panel's left edge and the note trigger's, in px. */
+  panelToTriggerGap: number
+}
+
+/**
+ * Measure the feedback row (and the open popover, when present) at the current
+ * viewport. The same reader serves the closed and open readings so the two
+ * sides differ only by whether the editor is open.
+ * @param page - the page under test.
+ * @param width - the viewport width already applied, recorded with the reading.
+ * @param editorOpen - true to also read the popover's relations; throws if it is absent.
+ * @returns the stop's relations.
+ */
+function measurePopover(page: Page, width: number, editorOpen: boolean): Promise<PopoverMetrics> {
+  return page.evaluate(({ viewportWidth, open }) => {
+    const rated = document.querySelector<HTMLElement>('button[aria-label="Remove rating"]')
+    if (rated === null) throw new Error('no rated feedback control in the DOM')
+    const row = rated.parentElement?.closest<HTMLElement>('div[class*="actions"]') ?? null
+    if (row === null) throw new Error('the IconActions row is not an ancestor of the feedback control')
+    const trigger = row.querySelector<HTMLElement>('button[aria-haspopup="dialog"]')
+    if (trigger === null) throw new Error('the note trigger is not in the row')
+
+    /**
+     * The real flex items of the row. A slot contributor (the feedback strip)
+     * arrives as a `display: contents` wrapper (the `assistant-actions` slot
+     * renders inside a transparent `data-slot` div), which reports an all-zero
+     * rect; a zero box would be miscounted as a phantom flex line. The actual
+     * items are the boxes inside it.
+     * @param element - the row whose items to read.
+     * @returns the real flex-item boxes, in flex/DOM order.
+     */
+    const flexItemBoxes = (element: HTMLElement): DOMRect[] => {
+      const boxes: DOMRect[] = []
+      for (const child of Array.from(element.children)) {
+        const el = child as HTMLElement
+        const rect = el.getBoundingClientRect()
+        if (el.style.display === 'contents') {
+          boxes.push(...flexItemBoxes(el))
+        } else if (rect.height > 0 && rect.width > 0) {
+          boxes.push(rect)
+        }
+      }
+      return boxes
+    }
+    /**
+     * Group items into flex lines by overlapping vertical extent.
+     * @param boxes - the row items' boxes, in DOM order.
+     * @returns the number of distinct lines.
+     */
+    const countFlexLines = (boxes: DOMRect[]): number => {
+      const centres: number[] = []
+      for (const box of boxes) {
+        const centre = box.top + box.height / 2
+        if (!centres.some(known => Math.abs(known - centre) <= box.height / 2)) centres.push(centre)
+      }
+      return centres.length
+    }
+
+    const column = row.closest<HTMLElement>('[data-conversation-scroll]')
+    const columnRight = (column?.getBoundingClientRect().left ?? 0) + (column?.clientWidth ?? 0)
+    const itemRects = flexItemBoxes(row)
+    // A half-pixel tolerance: subpixel layout puts a contained edge a fraction
+    // over the boundary on some device scale factors.
+    const itemsOutsideColumn = itemRects.filter(box => box.right > columnRight + 0.5).length
+    // The editor is a portal, so the row measures identically whether the
+    // editor is open or not; the closed/open fields differ by call so the sweep
+    // can assert a zero delta on them.
+    const overflow = row.scrollWidth - row.clientWidth
+
+    let builder: {
+      panelOutsideColumn: boolean
+      panelWithinViewport: boolean
+      panelToTriggerGap: number
+    }
+    if (!open) {
+      builder = { panelOutsideColumn: true, panelWithinViewport: true, panelToTriggerGap: 0 }
+    } else {
+      const panel = document.body.querySelector<HTMLElement>('[role="dialog"]')
+      if (panel === null) throw new Error('the note popover is not open')
+      const panelBox = panel.getBoundingClientRect()
+      const triggerBox = trigger.getBoundingClientRect()
+      const vw = window.innerWidth
+      const vh = window.innerHeight
+      builder = {
+        // The panel portals out of the column, so the clip cannot reach it.
+        panelOutsideColumn: column === null ? true : !column.contains(panel),
+        panelWithinViewport:
+          panelBox.left >= -0.5
+          && panelBox.right <= vw + 0.5
+          && panelBox.top >= -0.5
+          && panelBox.bottom <= vh + 0.5,
+        // The panel is fixed from the trigger's left, so a zero gap says it is
+        // anchored; a clamp can only widen it.
+        panelToTriggerGap: Math.abs(panelBox.left - triggerBox.left),
+      }
+    }
+
+    return {
+      width: viewportWidth,
+      rowOverflowClosed: overflow,
+      rowOverflowOpen: overflow,
+      rowLines: countFlexLines(itemRects),
+      itemsOutsideColumnClosed: itemsOutsideColumn,
+      itemsOutsideColumnOpen: itemsOutsideColumn,
+      ...builder,
+    }
+  }, { viewportWidth: width, open: editorOpen })
+}
+
+/**
+ * Render the golden body: one line per stop, relations and counts only. The
+ * row-overflow and outside-column readings are deltas (open minus closed) so
+ * the golden records that opening the editor leaves the row untouched, not an
+ * absolute count that many unrelated controls could move.
+ * @param stops - the measured stops, in sweep order.
+ * @returns the golden body, without a trailing newline.
+ */
+function renderGeometry(stops: PopoverMetrics[]): string {
+  return [
+    '# Assistant actions row with the feedback note popover open',
+    '',
+    '| viewport | row overflow delta | row lines | items-outside delta '
+      + '| panel outside the column | panel within the viewport | panel-to-trigger gap |',
+    '| --- | --- | --- | --- | --- | --- | --- |',
+    ...stops.map(stop => `| ${String(stop.width)}px | ${String(stop.rowOverflowOpen - stop.rowOverflowClosed)}px `
+      + `| ${String(stop.rowLines)} | ${String(stop.itemsOutsideColumnOpen - stop.itemsOutsideColumnClosed)} `
+      + `| ${String(stop.panelOutsideColumn)} | ${String(stop.panelWithinViewport)} `
+      + `| ${String(stop.panelToTriggerGap)}px |`),
+  ].join('\n')
+}
+
+describe('web e2e: the feedback note editor floats above the column', () => {
+  let scaffold: WebScaffold
+  let browser: Browser
+  let page: Page
+  let tripwire: ReturnType<typeof watchConsole>
+
+  beforeAll(async () => {
+    scaffold = await launchWebScaffold({})
+    await seedSession(scaffold, await readFile(SEED, 'utf8'), SEED_ID)
+    browser = await chromium.launch()
+    page = await newEnglishPage(browser, 900)
+    tripwire = watchConsole(page)
+    await page.goto(scaffold.baseUrl, { waitUntil: 'load' })
+    await page.waitForSelector('[class*="frame"]', { timeout: 30_000 })
+  }, 180_000)
+
+  afterAll(async () => {
+    await browser?.close()
+    await scaffold?.close()
+  })
+
+  /**
+   * Open the seeded transcript. The first treeitem is the collapsible group
+   * row; the session itself is the row beneath it.
+   * @returns nothing.
+   */
+  async function openSeededSession(): Promise<void> {
+    const groupRow = page.locator('[role="treeitem"]').first()
+    await groupRow.waitFor({ timeout: 15_000 })
+    if (await groupRow.getAttribute('aria-expanded') !== 'true') await groupRow.click()
+    const sessionRow = page.locator('[role="treeitem"]').nth(1)
+    await sessionRow.waitFor({ timeout: 15_000 })
+    await sessionRow.click()
+  }
+
+  /**
+   * Resize to a viewport and read the row once its width stops moving. The
+   * frame eases its column tracks, so reading straight after a resize can
+   * report the previous viewport's relation.
+   * @param width - viewport width to settle at.
+   * @param editorOpen - whether the note editor is currently open; reads the popover relations when so.
+   * @returns the row's (and popover's) readings at that width.
+   */
+  const settleAt = async (width: number, editorOpen: boolean): Promise<PopoverMetrics> => {
+    await page.setViewportSize({ width, height: 900 })
+    let previous = -1
+    await expect.poll(async () => {
+      const current = await page.evaluate(() =>
+        document.querySelector('[data-conversation-scroll]')?.clientWidth ?? -1)
+      const settled = current === previous
+      previous = current
+      return settled
+    }, { timeout: 10_000 }).toBe(true)
+    // The popover is JS-positioned from the trigger rect and re-places on
+    // resize/scroll, so once the column width stops moving we nudge it to the
+    // final layout; otherwise the panel can sit at a transient position from
+    // mid-resize and the anchor reading would be off.
+    await page.evaluate(() => window.dispatchEvent(new Event('resize')))
+    return measurePopover(page, width, editorOpen)
+  }
+
+  /**
+   * Rate a message, then for every stop read the row once with the note editor
+   * closed and once with it open, handing the SAME measured readings to both
+   * assertions so the golden and the assertions describe one measurement
+   * rather than two runs that could disagree.
+   * @returns the stops in {@link WIDTHS} order.
+   */
+  let swept: Promise<PopoverMetrics[]> | undefined
+  const sweep = (): Promise<PopoverMetrics[]> => {
+    swept ??= (async () => {
+      await openSeededSession()
+      await page.getByText('DONE', { exact: true }).waitFor({ timeout: 30_000 })
+      // The controller defers its list read to the first hover or focus, so the
+      // strip has to be touched before it can be rated.
+      const like = page.getByRole('button', { name: 'Good response' }).first()
+      await like.waitFor({ timeout: 30_000 })
+      await like.scrollIntoViewIfNeeded()
+      await like.hover()
+      await like.click()
+      await page.getByRole('button', { name: 'Remove rating' }).first()
+        .waitFor({ timeout: 15_000 })
+      const noteTrigger = page.getByRole('button', { name: 'Add a note' }).first()
+      const stops: PopoverMetrics[] = []
+      for (const width of WIDTHS) {
+        // Reset to the closed baseline at each stop before opening.
+        if (await noteTrigger.getAttribute('aria-expanded') === 'true') await noteTrigger.click()
+        const closed = await settleAt(width, false)
+        await page.getByRole('button', { name: 'Add a note' }).first().click()
+        await page.getByRole('dialog').waitFor({ timeout: 10_000 })
+        const open = await settleAt(width, true)
+        stops.push({
+          width,
+          rowOverflowClosed: closed.rowOverflowClosed,
+          rowOverflowOpen: open.rowOverflowOpen,
+          rowLines: open.rowLines,
+          itemsOutsideColumnClosed: closed.itemsOutsideColumnClosed,
+          itemsOutsideColumnOpen: open.itemsOutsideColumnOpen,
+          panelOutsideColumn: open.panelOutsideColumn,
+          panelWithinViewport: open.panelWithinViewport,
+          panelToTriggerGap: open.panelToTriggerGap,
+        })
+      }
+      return stops
+    })()
+    return swept
+  }
+
+  it('keeps the actions row untouched by the note popover, which stays in the viewport', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-message-feedback-layout'))
+    const stops = await sweep()
+    for (const stop of stops) {
+      // The popover lives outside the row, so opening it must not change the
+      // row at all. This is the vacuity guard of the whole redesign: an inline
+      // editor would widen or reflow the row, pushing the delta off zero.
+      expect(stop.rowOverflowOpen - stop.rowOverflowClosed, `viewport ${String(stop.width)}`).toBe(0)
+      expect(stop.itemsOutsideColumnOpen - stop.itemsOutsideColumnClosed, `viewport ${String(stop.width)}`).toBe(0)
+      // The row is one 28px line; the editor never forces a reflow.
+      expect(stop.rowLines, `viewport ${String(stop.width)}`).toBe(1)
+      // The panel escapes the column's overflow clip by living outside it.
+      expect(stop.panelOutsideColumn, `viewport ${String(stop.width)}`).toBe(true)
+      // The placement clamps the panel inside the viewport at every width.
+      expect(stop.panelWithinViewport, `viewport ${String(stop.width)}`).toBe(true)
+      // The panel stays anchored to its trigger rather than drifting off.
+      expect(stop.panelToTriggerGap, `viewport ${String(stop.width)}`).toBeLessThanOrEqual(4)
+    }
+    expect(tripwire.pageErrors).toEqual([])
+  }, 180_000)
+
+  it('matches the committed geometry golden', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-message-feedback-layout-golden'))
+    await compareOrRefreshGolden(GEOMETRY_EXPECTED, renderGeometry(await sweep()), MODE)
+  }, 180_000)
+
+  it('kept the console clean', () => {
+    expect(tripwire.pageErrors).toEqual([])
+    expect(tripwire.warnings).toEqual([])
+  })
+})

apps/web/tests/navigation-panes.e2e.ts

@@ -12,7 +12,7 @@ import { join } from 'node:path'
 import type { Browser, Page, Response } from 'playwright'
 import { chromium } from 'playwright'
 import { strFromU8, unzipSync } from 'fflate'
-import { afterAll, afterEach, beforeAll, beforeEach, describe, expect, it, onTestFailed } from 'vitest'
+import { afterAll, afterEach, beforeAll, beforeEach, describe, expect, it, onTestFailed, vi } from 'vitest'
 import { parseSessionLog } from '@deepseek-ai/dsh-llm-replay'
 import type { SessionEvent } from '@deepseek-ai/dsh-session'
 import {
@@ -415,8 +415,17 @@ describe('web e2e: navigation & panes over a rich seeded session', () => {
     // Read summaries are host-open file links; they also must not open details.
     const fileLink = page.locator('[data-variant="read"] button').first()
     await fileLink.waitFor({ timeout: 10_000 })
-    await fileLink.click()
-    await expect.poll(() => frame.getAttribute('data-details-collapsed'), { timeout: 5_000 }).toBe('true')
+    const openPath = vi.spyOn(scaffold.ctx.apiProxy.host, 'openPath')
+      .mockImplementation(async (request, _signal) => ({
+        rpcId: request.rpcId,
+        result: { ok: true, value: { opened: true as const } },
+      }))
+    try {
+      await fileLink.click()
+      await expect.poll(() => frame.getAttribute('data-details-collapsed'), { timeout: 5_000 }).toBe('true')
+    } finally {
+      openPath.mockRestore()
+    }
   }, 60_000)
 
   it.skipIf(MODE === 'record')('renders the bash row as a terminal card in the real browser', async () => {

apps/web/tests/seeded-history.e2e.ts

@@ -14,7 +14,7 @@ import { readFile, writeFile, mkdir } from 'node:fs/promises'
 import { fileURLToPath } from 'node:url'
 import type { Browser, Page } from 'playwright'
 import { chromium } from 'playwright'
-import { afterAll, beforeAll, describe, expect, it, onTestFailed } from 'vitest'
+import { afterAll, beforeAll, describe, expect, it, onTestFailed, vi } from 'vitest'
 import { createUserMessage } from '@deepseek-ai/dsh-llm'
 import type { ContentBlock, Message } from '@deepseek-ai/dsh-llm'
 import { deriveEventMessage, SessionId } from '@deepseek-ai/dsh-session'
@@ -33,6 +33,7 @@ const UI_EXPECTED = fileURLToPath(new URL('./snapshots/seeded-history/ui.expecte
 // Command-row goldens over the same conversation after direct host commands.
 const COMMAND_ROW_EXPECTED = fileURLToPath(new URL('./snapshots/seeded-history/command-row.expected.md', import.meta.url))
 const FEEDBACK_ROW_EXPECTED = fileURLToPath(new URL('./snapshots/seeded-history/feedback-row.expected.md', import.meta.url))
+const FILE_OPEN_FAILURE_EXPECTED = fileURLToPath(new URL('./snapshots/seeded-history/file-open-failure.expected.md', import.meta.url))
 const MODE = webSnapshotMode()
 const SEED_ID = 'seeded-history-web-e2e'
 
@@ -396,12 +397,57 @@ describe('web e2e: seeded history renders through cold resume', () => {
     await fileLink.waitFor({ timeout: 10_000 })
     const frame = page.locator('[style*="grid-template-columns"]').first()
     expect(await frame.getAttribute('data-details-collapsed')).toBe('true')
-    await fileLink.click()
-    await expect.poll(() => frame.getAttribute('data-details-collapsed'), { timeout: 5_000 }).toBe('true')
+    const openPath = vi.spyOn(scaffold.ctx.apiProxy.host, 'openPath')
+      .mockImplementation(async (request, _signal) => ({
+        rpcId: request.rpcId,
+        result: { ok: true, value: { opened: true as const } },
+      }))
+    try {
+      await fileLink.click()
+      await expect.poll(() => frame.getAttribute('data-details-collapsed'), { timeout: 5_000 }).toBe('true')
+    } finally {
+      openPath.mockRestore()
+    }
     // Path label survives from the recorded args (a.txt).
     await expect.poll(() => page.getByText('a.txt', { exact: false }).count(), { timeout: 5_000 }).toBeGreaterThan(0)
   })
 
+  it.skipIf(MODE === 'record')('a Host open refusal keeps the reason and retries the same path', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-seeded-file-open-failure'))
+    const fileLink = page.locator('[data-variant="read"] button').first()
+    await fileLink.waitFor({ timeout: 10_000 })
+    const openPath = vi.spyOn(scaffold.ctx.apiProxy.host, 'openPath')
+      .mockImplementation(async (request, _signal) => ({
+        rpcId: request.rpcId,
+        result: {
+          ok: false as const,
+          error: { code: 'internal', message: 'xdg-open is not available', details: {} },
+        },
+      }))
+    try {
+      await fileLink.click()
+      const dialog = page.getByRole('dialog', { name: 'Couldn’t open file' })
+      await dialog.waitFor({ timeout: 5_000 })
+      const snapshot = await captureStableAria(page, '[role="dialog"]', scaffold.workspaceCwd)
+      await compareOrRefreshGolden(FILE_OPEN_FAILURE_EXPECTED, snapshot, MODE)
+      await expect.poll(() => dialog.innerText(), { timeout: 5_000 })
+        .toContain('path open failed: xdg-open is not available')
+      await page.getByRole('button', { name: 'Retry' }).click()
+      await expect.poll(() => openPath.mock.calls.length, { timeout: 5_000 }).toBe(2)
+      expect(openPath.mock.calls[0]![0].payload).toEqual(openPath.mock.calls[1]![0].payload)
+      await page.getByRole('button', { name: 'Cancel' }).click()
+      await expect.poll(() => page.getByRole('dialog', { name: 'Couldn’t open file' }).count(), {
+        timeout: 5_000,
+      }).toBe(0)
+    } finally {
+      // Shared page: a leftover mask blocks later cases even when this one fails.
+      if (await page.getByRole('dialog', { name: 'Couldn’t open file' }).count() > 0) {
+        await page.keyboard.press('Escape')
+      }
+      openPath.mockRestore()
+    }
+  })
+
   it.skipIf(MODE === 'record')('expands the cold-resumed compact summary', async () => {
     onTestFailed(() => saveFailureShot(page, 'web-e2e-seeded-compaction'))
     const marker = page.getByRole('button', { name: /compact Compacted \d+ history items/ })
@@ -505,6 +551,6 @@ describe('web e2e: seeded history renders through cold resume', () => {
     // stream would have failed the turn loudly. Cleanliness pins the wire.
     expect(tripwire.pageErrors).toEqual([])
     expect(tripwire.warnings).toEqual([])
-    await assertFixtureInventory(SNAPSHOT_DIR, ['command-row.expected.md', 'feedback-row.expected.md', 'seed.jsonl', 'ui.expected.md'])
+    await assertFixtureInventory(SNAPSHOT_DIR, ['command-row.expected.md', 'feedback-row.expected.md', 'file-open-failure.expected.md', 'seed.jsonl', 'ui.expected.md'])
   })
 })

apps/web/tests/snapshots/home-path-tilde/workspace-hover.expected.txt

@@ -0,0 +1,2 @@
+hover=~/Documents/project
+copy=Copy: /home/fixture/Documents/project

apps/web/tests/snapshots/message-feedback-layout/geometry.expected.md

@@ -0,0 +1,10 @@
+# Assistant actions row with the feedback note popover open
+
+| viewport | row overflow delta | row lines | items-outside delta | panel outside the column | panel within the viewport | panel-to-trigger gap |
+| --- | --- | --- | --- | --- | --- | --- |
+| 1680px | 0px | 1 | 0 | true | true | 0px |
+| 1280px | 0px | 1 | 0 | true | true | 0px |
+| 1024px | 0px | 1 | 0 | true | true | 0px |
+| 900px | 0px | 1 | 0 | true | true | 0px |
+| 700px | 0px | 1 | 0 | true | true | 0px |
+| 600px | 0px | 1 | 0 | true | true | 0px |

apps/web/tests/snapshots/seeded-history/file-open-failure.expected.md

@@ -0,0 +1,7 @@
+- dialog "Couldn’t open file":
+  - heading "Couldn’t open file" [level=2]
+  - button "Close":
+    - img
+  - paragraph: "path open failed: xdg-open is not available"
+  - button "Cancel"
+  - button "Retry"

apps/web/tsconfig.json

@@ -60,6 +60,7 @@
     "tests/web-search-round.e2e.ts",
     "tests/message-actions.e2e.ts",
     "tests/message-feedback.e2e.ts",
+    "tests/message-feedback-layout.e2e.ts",
     "tests/markdown-images.e2e.ts",
     "tests/math-rendering.e2e.ts",
     "tests/markdown-cjk-strong.e2e.ts",

docs/config-catalog.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 docs/config-catalog.md
-config-catalog.md: 9c86924b53fac8df91bd75a8b1ba7b819d7b0194
-config-catalog.zh.md: bd462a78619027087469644cea9b579550ceef01
+config-catalog.md: 83594f198bf4b139d88349a15d57fe9a9de68a09
+config-catalog.zh.md: 175d2acbfb9d227e8f429c21c0d8f225b7e7ded4

docs/config-catalog.md

@@ -2118,7 +2118,7 @@ export interface Config {
 export type ClaudeCodePermissionMode = typeof CLAUDE_CODE_PERMISSION_MODES[number]
 ```
 
-Source: [`packages/subagent/subagent-claude-code/src/index.ts:37`](../packages/subagent/subagent-claude-code/src/index.ts)
+Source: [`packages/subagent/subagent-claude-code/src/index.ts:38`](../packages/subagent/subagent-claude-code/src/index.ts)
 
 <a id="deepseek-aidsh-subagent-codex"></a>
 
@@ -2149,7 +2149,7 @@ export type CodexPermissionMode =
   | 'dangerously-bypass-approvals-and-sandbox'
 ```
 
-Source: [`packages/subagent/subagent-codex/src/index.ts:35`](../packages/subagent/subagent-codex/src/index.ts)
+Source: [`packages/subagent/subagent-codex/src/index.ts:36`](../packages/subagent/subagent-codex/src/index.ts)
 
 <a id="deepseek-aidsh-subagent-dsh-sdk"></a>
 

docs/config-catalog.zh.md

@@ -2120,7 +2120,7 @@ export interface Config {
 export type ClaudeCodePermissionMode = typeof CLAUDE_CODE_PERMISSION_MODES[number]
 ```
 
-来源：[`packages/subagent/subagent-claude-code/src/index.ts:37`](../packages/subagent/subagent-claude-code/src/index.ts)
+来源：[`packages/subagent/subagent-claude-code/src/index.ts:38`](../packages/subagent/subagent-claude-code/src/index.ts)
 
 <a id="deepseek-aidsh-subagent-codex"></a>
 
@@ -2151,7 +2151,7 @@ export type CodexPermissionMode =
   | 'dangerously-bypass-approvals-and-sandbox'
 ```
 
-来源：[`packages/subagent/subagent-codex/src/index.ts:35`](../packages/subagent/subagent-codex/src/index.ts)
+来源：[`packages/subagent/subagent-codex/src/index.ts:36`](../packages/subagent/subagent-codex/src/index.ts)
 
 <a id="deepseek-aidsh-subagent-dsh-sdk"></a>
 

docs/module-graph.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 docs/module-graph.md
-module-graph.md: 5398515912c98962152d2b121b1b02d7a06097b9
-module-graph.zh.md: 7a398374603d66f8a3661a1d927b6418a99235a6
+module-graph.md: 38d3acc73175d504dadf22e4a5317e5f84ff5be8
+module-graph.zh.md: 76880c34aef07f877147a90f62dacd2b6877de41

docs/module-graph.md

@@ -1333,6 +1333,7 @@ flowchart TD
   pkg_client_ui_subagent --> pkg_subagent
   pkg_client_ui_subagent --> pkg_token_meter
   pkg_client_ui_tool --> pkg_api_remotes
+  pkg_client_ui_tool --> pkg_client_connection
   pkg_client_ui_tool --> pkg_client_locale
   pkg_client_ui_tool --> pkg_client_runtime
   pkg_client_ui_tool --> pkg_client_ui_conversation
@@ -1356,6 +1357,7 @@ flowchart TD
   pkg_client_ui_workflow_run --> pkg_session
   pkg_client_ui_workflow_run --> pkg_tool_workflow
   pkg_client_ui_workflow_run --> pkg_workflow
+  pkg_client_ui_workspace --> pkg_client_connection
   pkg_client_ui_workspace --> pkg_client_locale
   pkg_client_ui_workspace --> pkg_client_runtime
   pkg_client_ui_workspace --> pkg_client_ui_conversation
@@ -1625,11 +1627,11 @@ flowchart TD
 | [`client-ui-plan`](../packages/client/ui-plan) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants), [`plan-mode`](../packages/plan/plan-mode) |
 | [`client-ui-settings-general`](../packages/client/ui-settings-general) | `client` | [`api-remotes`](../packages/api/remotes), [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-settings`](../packages/client/ui-settings), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants), [`settings`](../packages/settings/settings) |
 | [`client-ui-subagent`](../packages/client/ui-subagent) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-input-trigger`](../packages/client/ui-input-trigger), [`invariants`](../packages/runtime-diagnostics/invariants), [`subagent`](../packages/subagent/subagent), [`token-meter`](../packages/llm/token-meter) |
-| [`client-ui-tool`](../packages/client/ui-tool) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
+| [`client-ui-tool`](../packages/client/ui-tool) | `client` | [`api-remotes`](../packages/api/remotes), [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-trajectory`](../packages/client/ui-trajectory) | `client` | [`agent`](../packages/core/agent), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`compaction`](../packages/compaction/compaction), [`invariants`](../packages/runtime-diagnostics/invariants), [`tools`](../packages/core/tools) |
 | [`client-ui-user-questions`](../packages/client/ui-user-questions) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-workflow-run`](../packages/client/ui-workflow-run) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants), [`session`](../packages/core/session), [`tool-workflow`](../packages/workflow/tool-workflow), [`workflow`](../packages/workflow/workflow) |
-| [`client-ui-workspace`](../packages/client/ui-workspace) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants) |
+| [`client-ui-workspace`](../packages/client/ui-workspace) | `client` | [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`session-log-export`](../packages/session-query/session-log-export) | `session-query` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-commands`](../packages/client/ui-commands), [`client-ui-conversation`](../packages/client/ui-conversation), [`commands`](../packages/interaction/commands), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-directory-picker-browse`](../packages/client/ui-directory-picker-browse) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-workspace`](../packages/client/ui-workspace), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-directory-picker-native`](../packages/client/ui-directory-picker-native) | `client` | [`client-runtime`](../packages/client/runtime), [`client-ui-workspace`](../packages/client/ui-workspace), [`invariants`](../packages/runtime-diagnostics/invariants) |

docs/module-graph.zh.md

@@ -1335,6 +1335,7 @@ flowchart TD
   pkg_client_ui_subagent --> pkg_subagent
   pkg_client_ui_subagent --> pkg_token_meter
   pkg_client_ui_tool --> pkg_api_remotes
+  pkg_client_ui_tool --> pkg_client_connection
   pkg_client_ui_tool --> pkg_client_locale
   pkg_client_ui_tool --> pkg_client_runtime
   pkg_client_ui_tool --> pkg_client_ui_conversation
@@ -1358,6 +1359,7 @@ flowchart TD
   pkg_client_ui_workflow_run --> pkg_session
   pkg_client_ui_workflow_run --> pkg_tool_workflow
   pkg_client_ui_workflow_run --> pkg_workflow
+  pkg_client_ui_workspace --> pkg_client_connection
   pkg_client_ui_workspace --> pkg_client_locale
   pkg_client_ui_workspace --> pkg_client_runtime
   pkg_client_ui_workspace --> pkg_client_ui_conversation
@@ -1627,11 +1629,11 @@ flowchart TD
 | [`client-ui-plan`](../packages/client/ui-plan) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants), [`plan-mode`](../packages/plan/plan-mode) |
 | [`client-ui-settings-general`](../packages/client/ui-settings-general) | `client` | [`api-remotes`](../packages/api/remotes), [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-settings`](../packages/client/ui-settings), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants), [`settings`](../packages/settings/settings) |
 | [`client-ui-subagent`](../packages/client/ui-subagent) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-input-trigger`](../packages/client/ui-input-trigger), [`invariants`](../packages/runtime-diagnostics/invariants), [`subagent`](../packages/subagent/subagent), [`token-meter`](../packages/llm/token-meter) |
-| [`client-ui-tool`](../packages/client/ui-tool) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
+| [`client-ui-tool`](../packages/client/ui-tool) | `client` | [`api-remotes`](../packages/api/remotes), [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-trajectory`](../packages/client/ui-trajectory) | `client` | [`agent`](../packages/core/agent), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`compaction`](../packages/compaction/compaction), [`invariants`](../packages/runtime-diagnostics/invariants), [`tools`](../packages/core/tools) |
 | [`client-ui-user-questions`](../packages/client/ui-user-questions) | `client` | [`api-remotes`](../packages/api/remotes), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-workflow-run`](../packages/client/ui-workflow-run) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`invariants`](../packages/runtime-diagnostics/invariants), [`session`](../packages/core/session), [`tool-workflow`](../packages/workflow/tool-workflow), [`workflow`](../packages/workflow/workflow) |
-| [`client-ui-workspace`](../packages/client/ui-workspace) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants) |
+| [`client-ui-workspace`](../packages/client/ui-workspace) | `client` | [`client-connection`](../packages/client/connection), [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-conversation`](../packages/client/ui-conversation), [`client-ui-sidebar`](../packages/client/ui-sidebar), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`session-log-export`](../packages/session-query/session-log-export) | `session-query` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-commands`](../packages/client/ui-commands), [`client-ui-conversation`](../packages/client/ui-conversation), [`commands`](../packages/interaction/commands), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-directory-picker-browse`](../packages/client/ui-directory-picker-browse) | `client` | [`client-locale`](../packages/client/locale), [`client-runtime`](../packages/client/runtime), [`client-ui-workspace`](../packages/client/ui-workspace), [`invariants`](../packages/runtime-diagnostics/invariants) |
 | [`client-ui-directory-picker-native`](../packages/client/ui-directory-picker-native) | `client` | [`client-runtime`](../packages/client/runtime), [`client-ui-workspace`](../packages/client/ui-workspace), [`invariants`](../packages/runtime-diagnostics/invariants) |

examples/acp-agent/tests/fixtures/subagent-result-diagnostic.ts

@@ -11,7 +11,28 @@ import { SessionId } from '@deepseek-ai/dsh-session'
 export const name = 'subagent-result-diagnostic'
 export const inject = ['subagents']
 
-const DIAGNOSTIC = 'Claude Code unattended decision (mode: dontAsk; request: tool permission; decision: denied): Claude Code denied the request before an interactive prompt'
+const RESULTS = [
+  {
+    id: '00000000-0000-4000-8000-0000000000d1',
+    diagnostic: 'Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)',
+    output: [{ type: 'text' as const, text: 'partial assistant text' }],
+  },
+  {
+    id: '00000000-0000-4000-8000-0000000000d2',
+    diagnostic: 'Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)',
+    output: [],
+  },
+  {
+    id: '00000000-0000-4000-8000-0000000000d3',
+    diagnostic: 'Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)',
+    output: [{ type: 'text' as const, text: 'partial assistant text' }],
+  },
+  {
+    id: '00000000-0000-4000-8000-0000000000d4',
+    diagnostic: 'Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)',
+    output: [],
+  },
+] as const
 
 class DiagnosticProvider implements SubagentProvider {
   readonly name = 'snapshot-diagnostic'
@@ -24,19 +45,16 @@ class DiagnosticProvider implements SubagentProvider {
       throw new Error('snapshot diagnostic provider start aborted')
     }
     const index = this.starts++
-    if (index > 1) {
-      throw new Error('snapshot diagnostic provider expected exactly two starts')
+    const fixture = RESULTS[index]
+    if (fixture === undefined) {
+      throw new Error('snapshot diagnostic provider expected exactly four starts')
     }
     return {
-      id: SessionId(index === 0
-        ? '00000000-0000-4000-8000-0000000000d1'
-        : '00000000-0000-4000-8000-0000000000d2'),
+      id: SessionId(fixture.id),
       localAgent: undefined,
       result: Promise.resolve({
-        output: index === 0
-          ? [{ type: 'text' as const, text: 'partial assistant text' }]
-          : [],
-        diagnostic: DIAGNOSTIC,
+        output: [...fixture.output],
+        diagnostic: fixture.diagnostic,
         stopReason: 'error' as const,
       }),
       dispose: async () => {},

examples/acp-agent/tests/snapshots/product-subagent-result-diagnostic/input.json

@@ -2,6 +2,6 @@
   "steps": [
     { "op": "initialize" },
     { "op": "newSession" },
-    { "op": "prompt", "text": "Use subagent_codex in the foreground exactly once; its result will fail with a diagnostic and partial output. Then use subagent_codex in the background exactly once and collect subagent-1 with job_output using wait true. After observing both failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools." }
+    { "op": "prompt", "text": "Observe four diagnostic failures with subagent_codex. First call it in the foreground for the Claude Code diagnostic, then in the background for the same Claude Code diagnostic and collect subagent-1 with job_output using wait true. Next call it in the foreground for the Codex diagnostic, then in the background for the same Codex diagnostic and collect subagent-2 with job_output using wait true. After all four failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools." }
   ]
 }

examples/acp-agent/tests/snapshots/product-subagent-result-diagnostic/replay.override.json

@@ -3,8 +3,8 @@
     "kind": "chunks",
     "chunks": [
       { "type": "block-start", "index": 0, "blockType": "tool-call" },
-      { "type": "tool-call-delta", "index": 0, "id": "call_diagnostic_foreground", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}" },
-      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_diagnostic_foreground", "name": "subagent_codex", "arguments": "{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}" } },
+      { "type": "tool-call-delta", "index": 0, "id": "call_claude_foreground", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_claude_foreground", "name": "subagent_codex", "arguments": "{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}" } },
       { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
       { "type": "finish", "reason": { "kind": "tool-calls" } }
     ]
@@ -13,8 +13,8 @@
     "kind": "chunks",
     "chunks": [
       { "type": "block-start", "index": 0, "blockType": "tool-call" },
-      { "type": "tool-call-delta", "index": 0, "id": "call_diagnostic_background", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}" },
-      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_diagnostic_background", "name": "subagent_codex", "arguments": "{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}" } },
+      { "type": "tool-call-delta", "index": 0, "id": "call_claude_background", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_claude_background", "name": "subagent_codex", "arguments": "{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}" } },
       { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
       { "type": "finish", "reason": { "kind": "tool-calls" } }
     ]
@@ -23,8 +23,38 @@
     "kind": "chunks",
     "chunks": [
       { "type": "block-start", "index": 0, "blockType": "tool-call" },
-      { "type": "tool-call-delta", "index": 0, "id": "call_diagnostic_output", "name": "job_output", "argumentsDelta": "{\"job_id\":\"subagent-1\",\"wait\":true}" },
-      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_diagnostic_output", "name": "job_output", "arguments": "{\"job_id\":\"subagent-1\",\"wait\":true}" } },
+      { "type": "tool-call-delta", "index": 0, "id": "call_claude_output", "name": "job_output", "argumentsDelta": "{\"job_id\":\"subagent-1\",\"wait\":true}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_claude_output", "name": "job_output", "arguments": "{\"job_id\":\"subagent-1\",\"wait\":true}" } },
+      { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
+      { "type": "finish", "reason": { "kind": "tool-calls" } }
+    ]
+  },
+  {
+    "kind": "chunks",
+    "chunks": [
+      { "type": "block-start", "index": 0, "blockType": "tool-call" },
+      { "type": "tool-call-delta", "index": 0, "id": "call_codex_foreground", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_codex_foreground", "name": "subagent_codex", "arguments": "{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}" } },
+      { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
+      { "type": "finish", "reason": { "kind": "tool-calls" } }
+    ]
+  },
+  {
+    "kind": "chunks",
+    "chunks": [
+      { "type": "block-start", "index": 0, "blockType": "tool-call" },
+      { "type": "tool-call-delta", "index": 0, "id": "call_codex_background", "name": "subagent_codex", "argumentsDelta": "{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_codex_background", "name": "subagent_codex", "arguments": "{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}" } },
+      { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
+      { "type": "finish", "reason": { "kind": "tool-calls" } }
+    ]
+  },
+  {
+    "kind": "chunks",
+    "chunks": [
+      { "type": "block-start", "index": 0, "blockType": "tool-call" },
+      { "type": "tool-call-delta", "index": 0, "id": "call_codex_output", "name": "job_output", "argumentsDelta": "{\"job_id\":\"subagent-2\",\"wait\":true}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_codex_output", "name": "job_output", "arguments": "{\"job_id\":\"subagent-2\",\"wait\":true}" } },
       { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
       { "type": "finish", "reason": { "kind": "tool-calls" } }
     ]

examples/acp-agent/tests/snapshots/product-subagent-result-diagnostic/session.jsonl

@@ -1,51 +1,84 @@
 {"type":"session","version":0,"id":"539aa64c-7f37-40ff-abd8-ed45b717be1b","createdAt":1783600629539,"cwd":"{{cwd}}","delegationDepth":0}
-{"type":"agent/inbox/spliced","seq":0,"time":1785498761270,"data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Use subagent_codex in the foreground exactly once; its result will fail with a diagnostic and partial output. Then use subagent_codex in the background exactly once and collect subagent-1 with job_output using wait true. After observing both failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools."}],"source":{"kind":"user"},"role":"user","id":"b8004c02-9892-40a7-b7a4-28f04879082c"}]}}
+{"type":"agent/inbox/spliced","seq":0,"time":1785498761270,"data":{"target":"next-turn","start":0,"inserted":[{"content":[{"type":"text","text":"Observe four diagnostic failures with subagent_codex. First call it in the foreground for the Claude Code diagnostic, then in the background for the same Claude Code diagnostic and collect subagent-1 with job_output using wait true. Next call it in the foreground for the Codex diagnostic, then in the background for the same Codex diagnostic and collect subagent-2 with job_output using wait true. After all four failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools."}],"source":{"kind":"user"},"role":"user","id":"eb9f20a0-9eac-480c-9904-71a1ffbb742a"}]}}
 {"type":"turn/start","seq":1,"time":1785821359466,"data":{"turn":1}}
 {"type":"agent/inbox/spliced","seq":2,"time":1785821359466,"data":{"target":"next-turn","start":0,"removedCount":1,"inserted":[]}}
 {"type":"step/start","seq":3,"time":1783600629542,"data":{"turn":1,"step":1}}
-{"type":"user/message","seq":4,"time":1785498761313,"data":{"content":[{"type":"text","text":"Use subagent_codex in the foreground exactly once; its result will fail with a diagnostic and partial output. Then use subagent_codex in the background exactly once and collect subagent-1 with job_output using wait true. After observing both failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools."}],"source":{"kind":"user"},"role":"user","id":"b8004c02-9892-40a7-b7a4-28f04879082c"},"surfaceOp":"append"}
+{"type":"user/message","seq":4,"time":1785498761313,"data":{"content":[{"type":"text","text":"Observe four diagnostic failures with subagent_codex. First call it in the foreground for the Claude Code diagnostic, then in the background for the same Claude Code diagnostic and collect subagent-1 with job_output using wait true. Next call it in the foreground for the Codex diagnostic, then in the background for the same Codex diagnostic and collect subagent-2 with job_output using wait true. After all four failures, reply with exactly PARENT_OBSERVED_DIAGNOSTICS. Do not call any other tools."}],"source":{"kind":"user"},"role":"user","id":"eb9f20a0-9eac-480c-9904-71a1ffbb742a"},"surfaceOp":"append"}
 {"type":"user/message","seq":5,"time":1785730415287,"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`)."}],"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`)."}]},"role":"user","id":"4b8d9730-0b7b-4e14-8a30-3d852f808f0e"},"surfaceOp":"append"}
-{"type":"session/title","seq":6,"time":1785730415287,"data":{"title":"Use subagent_codex in the foreground","messageSeqs":[4],"source":{"kind":"fallback"}}}
+{"type":"session/title","seq":6,"time":1785730415287,"data":{"title":"Observe four diagnostic failures with","messageSeqs":[4],"source":{"kind":"fallback"}}}
 {"type":"request/header","seq":7,"time":1785498761318,"data":{"header":{"config":{"provider":"deepseek-official","model":"deepseek-v4-pro"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"request/context","seq":8,"time":1785730415288,"data":{"provider":"deepseek-official","model":"deepseek-v4-pro"}}
 {"type":"assistant/chunk","seq":9,"time":1783600630822,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
-{"type":"assistant/chunk","seq":10,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_diagnostic_foreground","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}"}}}
-{"type":"assistant/chunk","seq":11,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_diagnostic_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}"}}}}
+{"type":"assistant/chunk","seq":10,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_claude_foreground","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}"}}}
+{"type":"assistant/chunk","seq":11,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_claude_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}"}}}}
 {"type":"assistant/chunk","seq":12,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":13,"time":1783600630852,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":14,"time":1786781990608,"data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_diagnostic_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"92e33995-2f02-4ad5-aec1-9df82cf4d583"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"}
-{"type":"tool/call","seq":15,"time":1786781990608,"data":{"turn":1,"step":1,"callId":"call_diagnostic_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe foreground diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":false}"}}
-{"type":"tool/result","seq":16,"time":1786781990613,"data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_diagnostic_foreground"},"content":[{"type":"tool-result","toolCallId":"call_diagnostic_foreground","content":[{"type":"text","text":"Error: subagent run failed\nDiagnostic: Claude Code unattended decision (mode: dontAsk; request: tool permission; decision: denied): Claude Code denied the request before an interactive prompt\nPartial output before the run ended:\npartial assistant text"}],"isError":true}],"role":"user","id":"4e84e7b3-40c1-488e-b119-45e8bd7ce448"}},"sourceEventSeqs":[15],"surfaceOp":"append"}
+{"type":"assistant/message","seq":14,"time":1786781990608,"data":{"turn":1,"step":1,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_claude_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"3cc2d0b5-97a5-4685-af60-ed7f7db8f69a"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[9,10,11,12,13],"surfaceOp":"append"}
+{"type":"tool/call","seq":15,"time":1786781990608,"data":{"turn":1,"step":1,"callId":"call_claude_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude foreground diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":false}"}}
+{"type":"tool/result","seq":16,"time":1786781990613,"data":{"turn":1,"step":1,"message":{"source":{"kind":"tool","callId":"call_claude_foreground"},"content":[{"type":"tool-result","toolCallId":"call_claude_foreground","content":[{"type":"text","text":"Error: subagent run failed\nDiagnostic: Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)\nPartial output before the run ended:\npartial assistant text"}],"isError":true}],"role":"user","id":"8743817e-158e-45cb-88d9-a695b2653eca"}},"sourceEventSeqs":[15],"surfaceOp":"append"}
 {"type":"step/end","seq":17,"time":1786781990613,"data":{"turn":1,"step":1}}
 {"type":"step/start","seq":18,"time":1786781990618,"data":{"turn":1,"step":2}}
 {"type":"assistant/chunk","seq":19,"time":1783600630926,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
-{"type":"assistant/chunk","seq":20,"time":1783600630926,"data":{"turn":1,"step":2,"chunk":{"type":"tool-call-delta","index":0,"id":"call_diagnostic_background","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}"}}}
-{"type":"assistant/chunk","seq":21,"time":1783600630926,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_diagnostic_background","name":"subagent_codex","arguments":"{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}"}}}}
+{"type":"assistant/chunk","seq":20,"time":1783600630926,"data":{"turn":1,"step":2,"chunk":{"type":"tool-call-delta","index":0,"id":"call_claude_background","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}"}}}
+{"type":"assistant/chunk","seq":21,"time":1783600630926,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_claude_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}"}}}}
 {"type":"assistant/chunk","seq":22,"time":1783600630944,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":23,"time":1783600630944,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":24,"time":1786781990622,"data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_diagnostic_background","name":"subagent_codex","arguments":"{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"2fb444e2-7a52-4963-988e-b1ecbc3744d5"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"}
-{"type":"tool/call","seq":25,"time":1786781990623,"data":{"turn":1,"step":2,"callId":"call_diagnostic_background","name":"subagent_codex","arguments":"{\"description\":\"Observe background diagnostic\",\"prompt\":\"Return the diagnostic failure.\",\"run_in_background\":true}"}}
-{"type":"agent/inbox/spliced","seq":26,"time":1786781990627,"data":{"target":"next-step","start":0,"inserted":[{"content":[{"type":"text","text":"background job subagent-1 (subagent: Observe background diagnostic) finished [status: failed, error; diagnostic: Claude Code unattended decision (mode: dontAsk; request: tool permission; decision: denied): Claude Code denied the request before an interactive prompt]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe background diagnostic [status: failed, error; diagnostic: Claude Code unattended decision (mode: dontA…"},"role":"user","id":"de606545-e637-4d9a-ba17-4c722a7331fd"}]}}
-{"type":"tool/result","seq":27,"time":1786781990627,"data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"call_diagnostic_background"},"content":[{"type":"tool-result","toolCallId":"call_diagnostic_background","content":[{"type":"text","text":"started background subagent job subagent-1"}],"isError":false}],"role":"user","id":"3377f724-b4a7-4ce1-bed7-774f174917d6"}},"sourceEventSeqs":[25],"surfaceOp":"append"}
+{"type":"assistant/message","seq":24,"time":1786781990622,"data":{"turn":1,"step":2,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_claude_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"b504312a-1dc5-46ce-87a5-12a5817511b9"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"}
+{"type":"tool/call","seq":25,"time":1786781990623,"data":{"turn":1,"step":2,"callId":"call_claude_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Claude background diagnostic\",\"prompt\":\"Return the Claude diagnostic failure.\",\"run_in_background\":true}"}}
+{"type":"agent/inbox/spliced","seq":26,"time":1786781990627,"data":{"target":"next-step","start":0,"inserted":[{"content":[{"type":"text","text":"background job subagent-1 (subagent: Observe Claude background diagnostic) finished [status: failed, error; diagnostic: Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe Claude background diagnostic [status: failed, error; diagnostic: Product subagent failure (product: Cl…"},"role":"user","id":"0fdb9ddf-1657-4455-9941-e6a9daa8ae4a"}]}}
+{"type":"tool/result","seq":27,"time":1786781990627,"data":{"turn":1,"step":2,"message":{"source":{"kind":"tool","callId":"call_claude_background"},"content":[{"type":"tool-result","toolCallId":"call_claude_background","content":[{"type":"text","text":"started background subagent job subagent-1"}],"isError":false}],"role":"user","id":"fe60646b-0551-4703-aa03-c8cb5460d356"}},"sourceEventSeqs":[25],"surfaceOp":"append"}
 {"type":"step/end","seq":28,"time":1786781990627,"data":{"turn":1,"step":2}}
 {"type":"agent/inbox/spliced","seq":29,"time":1786781990627,"data":{"target":"next-step","start":0,"removedCount":1,"inserted":[]}}
 {"type":"step/start","seq":30,"time":1786781990632,"data":{"turn":1,"step":3}}
-{"type":"user/message","seq":31,"time":1786781990632,"data":{"content":[{"type":"text","text":"background job subagent-1 (subagent: Observe background diagnostic) finished [status: failed, error; diagnostic: Claude Code unattended decision (mode: dontAsk; request: tool permission; decision: denied): Claude Code denied the request before an interactive prompt]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe background diagnostic [status: failed, error; diagnostic: Claude Code unattended decision (mode: dontA…"},"role":"user","id":"de606545-e637-4d9a-ba17-4c722a7331fd"},"surfaceOp":"append"}
+{"type":"user/message","seq":31,"time":1786781990632,"data":{"content":[{"type":"text","text":"background job subagent-1 (subagent: Observe Claude background diagnostic) finished [status: failed, error; diagnostic: Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe Claude background diagnostic [status: failed, error; diagnostic: Product subagent failure (product: Cl…"},"role":"user","id":"0fdb9ddf-1657-4455-9941-e6a9daa8ae4a"},"surfaceOp":"append"}
 {"type":"assistant/chunk","seq":32,"time":1783600631009,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
-{"type":"assistant/chunk","seq":33,"time":1783600631009,"data":{"turn":1,"step":3,"chunk":{"type":"tool-call-delta","index":0,"id":"call_diagnostic_output","name":"job_output","argumentsDelta":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}}
-{"type":"assistant/chunk","seq":34,"time":1783600631009,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_diagnostic_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}}}
+{"type":"assistant/chunk","seq":33,"time":1783600631009,"data":{"turn":1,"step":3,"chunk":{"type":"tool-call-delta","index":0,"id":"call_claude_output","name":"job_output","argumentsDelta":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}}
+{"type":"assistant/chunk","seq":34,"time":1783600631009,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_claude_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}}}
 {"type":"assistant/chunk","seq":35,"time":1785498761338,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":36,"time":1785730415297,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":37,"time":1785730415298,"data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_diagnostic_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"f43f988b-bc08-4811-8671-8edc0613f0d0"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"}
-{"type":"tool/call","seq":38,"time":1786781990636,"data":{"turn":1,"step":3,"callId":"call_diagnostic_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}
-{"type":"tool/result","seq":39,"time":1786781990640,"data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"call_diagnostic_output"},"content":[{"type":"tool-result","toolCallId":"call_diagnostic_output","content":[{"type":"text","text":"(no new output)\n[status: failed, error; diagnostic: Claude Code unattended decision (mode: dontAsk; request: tool permission; decision: denied): Claude Code denied the request before an interactive prompt]"}],"isError":false}],"role":"user","id":"6785120f-ae46-48d0-9f3f-d6cd1e6fc5d7"}},"sourceEventSeqs":[38],"surfaceOp":"append"}
+{"type":"assistant/message","seq":37,"time":1785730415298,"data":{"turn":1,"step":3,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_claude_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"c48a520a-74ed-42ee-9d93-ee59899975b0"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[32,33,34,35,36],"surfaceOp":"append"}
+{"type":"tool/call","seq":38,"time":1786781990636,"data":{"turn":1,"step":3,"callId":"call_claude_output","name":"job_output","arguments":"{\"job_id\":\"subagent-1\",\"wait\":true}"}}
+{"type":"tool/result","seq":39,"time":1786781990640,"data":{"turn":1,"step":3,"message":{"source":{"kind":"tool","callId":"call_claude_output"},"content":[{"type":"tool-result","toolCallId":"call_claude_output","content":[{"type":"text","text":"(no new output)\n[status: failed, error; diagnostic: Product subagent failure (product: Claude Code; stage: query-run; category: error_max_budget_usd)]"}],"isError":false}],"role":"user","id":"45bc0705-7243-4173-a119-4c0655af8dc1"}},"sourceEventSeqs":[38],"surfaceOp":"append"}
 {"type":"step/end","seq":40,"time":1786781990640,"data":{"turn":1,"step":3}}
 {"type":"step/start","seq":41,"time":1786781990645,"data":{"turn":1,"step":4}}
-{"type":"assistant/chunk","seq":42,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
-{"type":"assistant/chunk","seq":43,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"text-delta","index":0,"text":"PARENT_OBSERVED_DIAGNOSTICS"}}}
-{"type":"assistant/chunk","seq":44,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"PARENT_OBSERVED_DIAGNOSTICS"}}}}
-{"type":"assistant/chunk","seq":45,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":2}}}}
-{"type":"assistant/chunk","seq":46,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":47,"time":1786781990649,"data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"text","text":"PARENT_OBSERVED_DIAGNOSTICS"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"49b868e8-2608-47e0-aaf8-b308ffe8194d"},"usage":{"inputTokens":10,"outputTokens":2}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"}
-{"type":"step/end","seq":48,"time":1786781990650,"data":{"turn":1,"step":4}}
-{"type":"turn/end","seq":49,"time":1786781990650,"data":{"turn":1,"reason":{"kind":"completed"}}}
+{"type":"assistant/chunk","seq":42,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":43,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"tool-call-delta","index":0,"id":"call_codex_foreground","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}"}}}
+{"type":"assistant/chunk","seq":44,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_codex_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}"}}}}
+{"type":"assistant/chunk","seq":45,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":46,"time":1786781990649,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":47,"time":1786781990649,"data":{"turn":1,"step":4,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_codex_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"89ab3728-fc3f-4825-97e5-383d46568d8c"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[42,43,44,45,46],"surfaceOp":"append"}
+{"type":"tool/call","seq":48,"time":1786994591759,"data":{"turn":1,"step":4,"callId":"call_codex_foreground","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex foreground diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":false}"}}
+{"type":"tool/result","seq":49,"time":1786994591762,"data":{"turn":1,"step":4,"message":{"source":{"kind":"tool","callId":"call_codex_foreground"},"content":[{"type":"tool-result","toolCallId":"call_codex_foreground","content":[{"type":"text","text":"Error: subagent run failed\nDiagnostic: Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)\nPartial output before the run ended:\npartial assistant text"}],"isError":true}],"role":"user","id":"0a8fd87c-eacb-457b-a5ad-29dd88f599aa"}},"sourceEventSeqs":[48],"surfaceOp":"append"}
+{"type":"step/end","seq":50,"time":1786994591762,"data":{"turn":1,"step":4}}
+{"type":"step/start","seq":51,"time":1786994591767,"data":{"turn":1,"step":5}}
+{"type":"assistant/chunk","seq":52,"time":1786994591771,"data":{"turn":1,"step":5,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":53,"time":1786994591771,"data":{"turn":1,"step":5,"chunk":{"type":"tool-call-delta","index":0,"id":"call_codex_background","name":"subagent_codex","argumentsDelta":"{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}"}}}
+{"type":"assistant/chunk","seq":54,"time":1786994591771,"data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_codex_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}"}}}}
+{"type":"assistant/chunk","seq":55,"time":1786994591771,"data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":56,"time":1786994591771,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":57,"time":1786994591771,"data":{"turn":1,"step":5,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_codex_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"996da601-acb8-49c9-8dd7-e60a88a8f1a2"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[52,53,54,55,56],"surfaceOp":"append"}
+{"type":"tool/call","seq":58,"time":1786994591772,"data":{"turn":1,"step":5,"callId":"call_codex_background","name":"subagent_codex","arguments":"{\"description\":\"Observe Codex background diagnostic\",\"prompt\":\"Return the Codex diagnostic failure.\",\"run_in_background\":true}"}}
+{"type":"agent/inbox/spliced","seq":59,"time":1786994591775,"data":{"target":"next-step","start":0,"inserted":[{"content":[{"type":"text","text":"background job subagent-2 (subagent: Observe Codex background diagnostic) finished [status: failed, error; diagnostic: Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe Codex background diagnostic [status: failed, error; diagnostic: Product subagent failure (product: Cod…"},"role":"user","id":"5f1d4517-50d4-48ef-8acc-8f9361ecb185"}]}}
+{"type":"tool/result","seq":60,"time":1786994591775,"data":{"turn":1,"step":5,"message":{"source":{"kind":"tool","callId":"call_codex_background"},"content":[{"type":"tool-result","toolCallId":"call_codex_background","content":[{"type":"text","text":"started background subagent job subagent-2"}],"isError":false}],"role":"user","id":"a8ba8362-275b-4bca-8b88-d1ef84d325a3"}},"sourceEventSeqs":[58],"surfaceOp":"append"}
+{"type":"step/end","seq":61,"time":1786994591776,"data":{"turn":1,"step":5}}
+{"type":"agent/inbox/spliced","seq":62,"time":1786994591776,"data":{"target":"next-step","start":0,"removedCount":1,"inserted":[]}}
+{"type":"step/start","seq":63,"time":1786994591781,"data":{"turn":1,"step":6}}
+{"type":"user/message","seq":64,"time":1786994591781,"data":{"content":[{"type":"text","text":"background job subagent-2 (subagent: Observe Codex background diagnostic) finished [status: failed, error; diagnostic: Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)]. Read its output with job_output."}],"source":{"kind":"plugin","plugin":"tool-jobs","form":"notice","summary":"subagent Observe Codex background diagnostic [status: failed, error; diagnostic: Product subagent failure (product: Cod…"},"role":"user","id":"5f1d4517-50d4-48ef-8acc-8f9361ecb185"},"surfaceOp":"append"}
+{"type":"assistant/chunk","seq":65,"time":1786994591788,"data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":66,"time":1786994591788,"data":{"turn":1,"step":6,"chunk":{"type":"tool-call-delta","index":0,"id":"call_codex_output","name":"job_output","argumentsDelta":"{\"job_id\":\"subagent-2\",\"wait\":true}"}}}
+{"type":"assistant/chunk","seq":67,"time":1786994591788,"data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_codex_output","name":"job_output","arguments":"{\"job_id\":\"subagent-2\",\"wait\":true}"}}}}
+{"type":"assistant/chunk","seq":68,"time":1786994591789,"data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":69,"time":1786994591789,"data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":70,"time":1786994591789,"data":{"turn":1,"step":6,"message":{"role":"assistant","content":[{"type":"tool-call","id":"call_codex_output","name":"job_output","arguments":"{\"job_id\":\"subagent-2\",\"wait\":true}"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"cfc1726c-5d9e-486a-aa0f-057219e16dfd"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[65,66,67,68,69],"surfaceOp":"append"}
+{"type":"tool/call","seq":71,"time":1786994591789,"data":{"turn":1,"step":6,"callId":"call_codex_output","name":"job_output","arguments":"{\"job_id\":\"subagent-2\",\"wait\":true}"}}
+{"type":"tool/result","seq":72,"time":1786994591797,"data":{"turn":1,"step":6,"message":{"source":{"kind":"tool","callId":"call_codex_output"},"content":[{"type":"tool-result","toolCallId":"call_codex_output","content":[{"type":"text","text":"(no new output)\n[status: failed, error; diagnostic: Product subagent failure (product: Codex; stage: turn; category: httpConnectionFailed; HTTP status: 503)]"}],"isError":false}],"role":"user","id":"9671e5ee-f443-4548-8fd2-b0b76f00b629"}},"sourceEventSeqs":[71],"surfaceOp":"append"}
+{"type":"step/end","seq":73,"time":1786994591797,"data":{"turn":1,"step":6}}
+{"type":"step/start","seq":74,"time":1786994591802,"data":{"turn":1,"step":7}}
+{"type":"assistant/chunk","seq":75,"time":1786994591806,"data":{"turn":1,"step":7,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":76,"time":1786994591806,"data":{"turn":1,"step":7,"chunk":{"type":"text-delta","index":0,"text":"PARENT_OBSERVED_DIAGNOSTICS"}}}
+{"type":"assistant/chunk","seq":77,"time":1786994591806,"data":{"turn":1,"step":7,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"PARENT_OBSERVED_DIAGNOSTICS"}}}}
+{"type":"assistant/chunk","seq":78,"time":1786994591806,"data":{"turn":1,"step":7,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":2}}}}
+{"type":"assistant/chunk","seq":79,"time":1786994591806,"data":{"turn":1,"step":7,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":80,"time":1786994591806,"data":{"turn":1,"step":7,"message":{"role":"assistant","content":[{"type":"text","text":"PARENT_OBSERVED_DIAGNOSTICS"}],"source":{"kind":"model","provider":"deepseek-official","model":"deepseek-v4-pro"},"id":"49b868e8-2608-47e0-aaf8-b308ffe8194d"},"usage":{"inputTokens":10,"outputTokens":2}},"sourceEventSeqs":[75,76,77,78,79],"surfaceOp":"append"}
+{"type":"step/end","seq":81,"time":1786994591806,"data":{"turn":1,"step":7}}
+{"type":"turn/end","seq":82,"time":1786994591807,"data":{"turn":1,"reason":{"kind":"completed"}}}

packages/client/connection/src/client/fixture.ts

@@ -1563,6 +1563,7 @@ function createFixtureWorld(options: FixtureOptions): FixtureWorld {
   // live under one workspace, whose account carries them in attach order.
   const wid = (raw: string): WorkspaceId => raw as WorkspaceId
   const fixtureEpoch = new Date(Date.now() - 300_000).toISOString()
+  const FIXTURE_HOME = '/home/fixture'
   const workspaces: WorkspaceView[] = options.empty ? [] : [{
     workspaceId: wid('fx-ws-fixture'),
     path: '/tmp/fixture',
@@ -1570,6 +1571,13 @@ function createFixtureWorld(options: FixtureOptions): FixtureWorld {
     sessionIds: [sid('fx-alpha'), sid('fx-beta'), sid('fx-gamma')],
     createdAt: fixtureEpoch,
     updatedAt: fixtureEpoch,
+  }, {
+    workspaceId: wid('fx-ws-home'),
+    path: `${FIXTURE_HOME}/Documents/project`,
+    title: 'project',
+    sessionIds: [],
+    createdAt: fixtureEpoch,
+    updatedAt: fixtureEpoch,
   }]
   let nextWorkspace = 1
   // Registry-global archive set mirroring the host: archived sessions keep
@@ -1580,7 +1588,6 @@ function createFixtureWorld(options: FixtureOptions): FixtureWorld {
   // deterministic content mirroring the design mock so assembled Web tests
   // and snapshots can walk it. Leaves are materialized lazily: a child listed
   // by its parent lists as empty until something is created inside it.
-  const FIXTURE_HOME = '/home/fixture'
   const directoryTree = new Map<string, string[]>([
     ['/', ['home']],
     ['/home', ['fixture']],
@@ -2565,7 +2572,7 @@ function createFixtureWorld(options: FixtureOptions): FixtureWorld {
     },
     host: {
       describe: request => ok(request, {
-        version: '0.0.0-fixture', cwd: '/tmp/fixture', attachedSessions, canOpenPath: true,
+        version: '0.0.0-fixture', cwd: '/tmp/fixture', attachedSessions, home: FIXTURE_HOME, canOpenPath: true,
       }),
       // Deterministic native pick: the keyless lanes drive the full
       // pick-then-adopt path without an OS chooser (design-mock content,

packages/client/connection/src/client/index.ts

@@ -62,7 +62,7 @@ export interface ConnectionHandle {
   readonly api: IApiClient
   /** Whether the current page authority is loopback; non-browser contexts default to true. */
   readonly isLoopback: boolean
-  /** Generation-scoped Host facts, including native path-open capability. */
+  /** Generation-scoped Host facts, including the account home and native path-open capability. */
   readonly hostDescription: HostDescriptionSource
   /** Generic logical RPC channels over the same Connection transport. */
   readonly rpc: ClientConnectionRpc

packages/client/connection/tests/connection.client.spec.ts

@@ -80,7 +80,7 @@ describe('connection lifecycle', () => {
     try {
       await vi.waitFor(() => { expect(describeCalls).toBe(2) }) // retried after backoff
       expect(connected).toBe(0) // never announced during the failed generation
-      gate.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }))
+      gate.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }))
       await vi.waitFor(() => { expect(connected).toBe(1) })
     } finally {
       controller.stop()
@@ -102,7 +102,7 @@ describe('connection lifecycle', () => {
           },
         })
       }
-      return Promise.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }))
+      return Promise.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }))
     }
     let connected = 0
     const warnSpy = vi.spyOn(console, 'warn').mockImplementation(() => undefined)
@@ -188,7 +188,7 @@ describe('connection lifecycle', () => {
       describeCalls++
       return describeCalls === 1
         ? firstDescribe.promise
-        : Promise.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }))
+        : Promise.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }))
     }
     const states: ConnectionState[] = []
     let connected = 0
@@ -201,7 +201,7 @@ describe('connection lifecycle', () => {
     try {
       await vi.waitFor(() => { expect(api.openMuxCount).toBe(1) })
       api.endStreams()
-      firstDescribe.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }))
+      firstDescribe.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }))
 
       await vi.waitFor(() => { expect(describeCalls).toBe(2) })
       await vi.waitFor(() => { expect(connected).toBe(1) })
@@ -283,7 +283,7 @@ describe('connection lifecycle', () => {
     controller.start()
     try {
       await vi.waitFor(() => { expect(describeCalls).toBe(3) })
-      gate.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }))
+      gate.resolve(ok({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }))
       await vi.waitFor(() => { expect(connected).toBe(1) })
       expect(states).toEqual(['reconnecting', 'connected']) // two failures, one reconnecting emission
     } finally {

packages/client/connection/tests/fake-api.client.ts

@@ -74,10 +74,11 @@ export class FakeApiClient implements IApiClient {
     version: string
     cwd: string
     attachedSessions: number
+    home: string
     canOpenPath: boolean
   }>> =
     () => Promise.resolve(ok({
-      version: '0-fake', cwd: '/f', attachedSessions: 0, canOpenPath: true,
+      version: '0-fake', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true,
     }))
   onPickDirectory: (payload: unknown) => Promise<RpcResponse<{ path: string | null }>> =
     () => Promise.resolve(ok({ path: null }))

packages/client/connection/tests/fixture.client.spec.ts

@@ -523,7 +523,9 @@ describe('createFixtureApi', () => {
   it('describe answers the fixture identity', async () => {
     const api = createFixtureApi()
     const response = await api.host.describe(req({}))
-    expect(response.result).toMatchObject({ ok: true, value: { version: '0.0.0-fixture', attachedSessions: 1 } })
+    expect(response.result).toMatchObject({
+      ok: true, value: { version: '0.0.0-fixture', attachedSessions: 1, home: '/home/fixture' },
+    })
     const empty = await createFixtureApi({ empty: true }).host.describe(req({}))
     expect(empty.result).toMatchObject({ ok: true, value: { attachedSessions: 0 } })
   })
@@ -548,10 +550,16 @@ describe('createFixtureApi', () => {
     const api = createFixtureApi()
     const listed = await api.workspace.list(req({}))
     if (!listed.result.ok) throw new Error('list failed')
-    expect(listed.result.value.items).toEqual([expect.objectContaining({
-      workspaceId: 'fx-ws-fixture', path: '/tmp/fixture', title: 'fixture',
-      sessionIds: ['fx-alpha', 'fx-beta', 'fx-gamma'],
-    })])
+    expect(listed.result.value.items).toEqual([
+      expect.objectContaining({
+        workspaceId: 'fx-ws-fixture', path: '/tmp/fixture', title: 'fixture',
+        sessionIds: ['fx-alpha', 'fx-beta', 'fx-gamma'],
+      }),
+      expect.objectContaining({
+        workspaceId: 'fx-ws-home', path: '/home/fixture/Documents/project', title: 'project',
+        sessionIds: [],
+      }),
+    ])
     // path collision → the existing entity comes back, created:false, no frame.
     const reused = await api.workspace.create(req({ path: '/tmp/fixture' }))
     if (!reused.result.ok) throw new Error('reuse failed')

packages/client/runtime/README.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 packages/client/runtime/README.md
-README.md: cad8a08f015a7115ebff76cfce4d3ab5eb7a175e
-README.zh.md: 262ab1eeb8ce2d09e801e2af790046d8cdc11e55
+README.md: 7d0e90809aaffb9503f2222e1f7426b13d72d7d5
+README.zh.md: d0a43c32b4861d02786aa0c42e1af72047c9032a

packages/client/runtime/README.md

@@ -26,6 +26,8 @@ Workspace and Session lists have independent monotone `pending` → `ready` base
 
 SlotRegistry gives the renderer separate bare observables for `useSessions` and `useWorkspaces`; ui-renderer creates the hooks. Workspace business state does not enter `SessionListState` or an entry store.
 
+`abbreviateHomePath` is the display-only POSIX home abbreviation used by Web Workspace hover cards and Tool summaries; a Windows drive or UNC path stays verbatim, and a missing, empty, or filesystem-root home leaves the path unchanged.
+
 `indexSubagentDescendants()` derives per-parent total and running descendant counts from the retained list mirror. It follows only uninterrupted `origin: 'subagent'` ancestry, so an ordinary fork starts a separate ownership subtree; cycles stop without throwing, and a missing parent remains a harmless key until its summary arrives.
 
 `SessionListState.jobsBySession` mirrors the Host's `session/jobs` frames last-wins, keyed by session and needing no Session instance. An emptied set is stored as an absent key, so absence and `[]` are one representation and consumers never test a sentinel. Two clears keep it from outliving its truth: `session/subscribed` drops the session's mirror, because a fresh generation sends a baseline only for a non-empty set and a retained list would survive as a phantom, and `host/session-removed` drops it again, because owner disposal removed the records on the mux stream while the removal frame rides the host stream, leaving the two with no relative order.

packages/client/runtime/README.zh.md

@@ -26,6 +26,8 @@ Workspace 和 Session 列表各自具有单调的 `pending` → `ready` 基线
 
 SlotRegistry 分别为 renderer 提供 `useSessions` 与 `useWorkspaces` 的裸 observable；ui-renderer 创建钩子。Workspace 业务状态不会进入 `SessionListState` 或条目 store。
 
+`abbreviateHomePath` 是 Web Workspace 悬停卡片与 Tool 摘要使用的仅展示 POSIX 家目录缩写；Windows 盘符或 UNC 路径保持原样，缺失、空或文件系统根的 home 不改写路径。
+
 `indexSubagentDescendants()` 从保留的列表镜像中派生每个 parent 的后代总数与运行中后代数。它只沿不间断的 `origin: 'subagent'` 祖先链追踪，因此普通 fork 会开启独立的归属子树；遇到环时，追踪会停止但不会抛出异常，缺失的 parent 则会保留为无害的键，直至其摘要到达。
 
 `SessionListState.jobsBySession` 按 last-wins 镜像宿主的 `session/jobs` 帧，以会话为键，不需要 Session 实例。被清空的集合存为缺失的键，因此「缺失」与 `[]` 是同一种表示，消费方永远不必检测哨兵值。两处清理让它不至于比它所反映的真相活得更久：`session/subscribed` 丢弃该会话的镜像，因为新一代只为非空集合发送 baseline，被留下的列表会变成幽灵；`host/session-removed` 再丢一次，因为 owner 销毁是在 mux 流上移除记录的，而移除帧走 host 流，两者没有相对顺序。

packages/client/runtime/src/client/index.ts

@@ -46,7 +46,7 @@ export type { SessionProvideChannelHost } from './sessions/provide.ts'
 export { createScope } from './agents/scope.ts'
 export type { AgentScopeHandle } from './agents/scope.ts'
 export { DirectoryBrowseError, WorkspaceCreateError, WorkspaceRuntime } from './workspaces/service.ts'
-export { resolveWorkspacePath } from './workspaces/path.ts'
+export { abbreviateHomePath, resolveWorkspacePath } from './workspaces/path.ts'
 // Contract only: the scope implementation and its Host transport belong to
 // dsh-client-ui-settings (see that package's settings-scope.ts).
 export type {

packages/client/runtime/src/client/workspaces/path.ts

@@ -5,9 +5,32 @@
  * @returns an absolute path when a workspace root is available, otherwise the original path.
  */
 export function resolveWorkspacePath(cwd: string | undefined, path: string): string {
-  if (path.startsWith('/') || /^[A-Za-z]:[/\\]/.test(path) || path.startsWith('\\\\')) return path
+  if (path.startsWith('/') || isWindowsStylePath(path)) return path
   if (cwd === undefined || cwd === '') return path
   const base = cwd.replace(/[/\\]+$/, '')
   const rel = path.replace(/^[/\\]+/, '')
   return `${base}/${rel}`
 }
+
+/** Drive-letter or UNC path; Web display must not rewrite these as `~`. */
+function isWindowsStylePath(value: string): boolean {
+  return /^[A-Za-z]:[/\\]/.test(value) || value.startsWith('\\\\')
+}
+
+/**
+ * Display-only POSIX home abbreviation. Windows drive and UNC paths stay
+ * verbatim, including when `home` itself is a Windows path. A missing, empty,
+ * or filesystem-root `home` leaves `path` unchanged so `/` cannot become `~`.
+ * @param path - absolute or already-short display path.
+ * @param home - host account home from `host.describe`; absent skips abbreviation.
+ * @returns `~` or `~/…` for the POSIX home and its descendants, otherwise `path`.
+ */
+export function abbreviateHomePath(path: string, home?: string): string {
+  if (home === undefined || home === '') return path
+  if (isWindowsStylePath(path) || isWindowsStylePath(home)) return path
+  const root = home.replace(/\/+$/, '')
+  if (root === '' || root === '/') return path
+  if (path.replace(/\/+$/, '') === root) return '~'
+  if (path.startsWith(`${root}/`)) return `~${path.slice(root.length)}`
+  return path
+}

packages/client/runtime/tests/client-apply.client.spec.ts

@@ -91,7 +91,7 @@ describe('runtime client apply', () => {
     expect(workspaces.list.getSnapshot().items[0]?.workspaceId).toBe('w-new')
     // Mux sink and onConnected route without throwing (manager semantics own the behavior).
     bench.sinks?.onMuxEnvelope?.({ rpcId: 'r2' as never, payload: { type: 'stream/error', message: 'x' } as never })
-    bench.sinks?.onConnected?.({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true })
+    bench.sinks?.onConnected?.({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true })
   })
 
   it('selects the recent Workspace once when the first baselines have no current session', async () => {
@@ -104,7 +104,7 @@ describe('runtime client apply', () => {
     }))
     bench.api.onList = () => Promise.resolve(ok({ items: [] }))
 
-    bench.sinks?.onConnected?.({ version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true })
+    bench.sinks?.onConnected?.({ version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true })
     await flushMicrotasks()
 
     const sessions = bench.ctx.get('sessions') as SessionRuntime

packages/client/runtime/tests/fake-api.client.ts

@@ -108,10 +108,11 @@ export class FakeApiClient implements IApiClient {
     version: string
     cwd: string
     attachedSessions: number
+    home: string
     canOpenPath: boolean
   }>> =
     () => Promise.resolve(ok({
-      version: '0-fake', cwd: '/f', attachedSessions: 0, canOpenPath: true,
+      version: '0-fake', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true,
     }))
   onPickDirectory: (payload: unknown) => Promise<RpcResponse<{ path: string | null }>> =
     () => Promise.resolve(ok({ path: null }))

packages/client/runtime/tests/path.client.spec.ts

@@ -0,0 +1,41 @@
+import { describe, expect, it } from 'vitest'
+import { abbreviateHomePath, resolveWorkspacePath } from '../src/client/workspaces/path.ts'
+
+describe('abbreviateHomePath', () => {
+  it('collapses a POSIX home and its descendants', () => {
+    expect(abbreviateHomePath('/Users/u', '/Users/u')).toBe('~')
+    expect(abbreviateHomePath('/Users/u/', '/Users/u')).toBe('~')
+    expect(abbreviateHomePath('/Users/u/Documents/project', '/Users/u')).toBe('~/Documents/project')
+    expect(abbreviateHomePath('/Users/u/Documents/project/', '/Users/u/')).toBe('~/Documents/project/')
+  })
+
+  it('keeps prefix-adjacent names and non-home paths', () => {
+    expect(abbreviateHomePath('/Users/u2/a.ts', '/Users/u')).toBe('/Users/u2/a.ts')
+    expect(abbreviateHomePath('/etc/hosts', '/Users/u')).toBe('/etc/hosts')
+    expect(abbreviateHomePath('src/a.ts', '/Users/u')).toBe('src/a.ts')
+    expect(abbreviateHomePath('~/already', '/Users/u')).toBe('~/already')
+  })
+
+  it('does not abbreviate when home is missing, empty, or the filesystem root', () => {
+    expect(abbreviateHomePath('/Users/u/a.ts')).toBe('/Users/u/a.ts')
+    expect(abbreviateHomePath('/Users/u/a.ts', '')).toBe('/Users/u/a.ts')
+    expect(abbreviateHomePath('/etc/hosts', '/')).toBe('/etc/hosts')
+    expect(abbreviateHomePath('/etc/hosts', '///')).toBe('/etc/hosts')
+  })
+
+  it('leaves Windows drive and UNC paths verbatim', () => {
+    expect(abbreviateHomePath('C:\\Users\\u\\project', 'C:\\Users\\u')).toBe('C:\\Users\\u\\project')
+    expect(abbreviateHomePath('C:/Users/u/project', '/Users/u')).toBe('C:/Users/u/project')
+    expect(abbreviateHomePath('/Users/u/project', 'C:\\Users\\u')).toBe('/Users/u/project')
+    expect(abbreviateHomePath('\\\\server\\share\\u', '\\\\server\\share\\u')).toBe('\\\\server\\share\\u')
+  })
+})
+
+describe('resolveWorkspacePath', () => {
+  it('joins a relative path under cwd and passes absolute paths through', () => {
+    expect(resolveWorkspacePath('/w', 'src/a.ts')).toBe('/w/src/a.ts')
+    expect(resolveWorkspacePath('/w/', '/abs/a.ts')).toBe('/abs/a.ts')
+    expect(resolveWorkspacePath(undefined, 'src/a.ts')).toBe('src/a.ts')
+    expect(resolveWorkspacePath('/w', 'C:\\x\\a.ts')).toBe('C:\\x\\a.ts')
+  })
+})

packages/client/runtime/tests/wire-events.client.spec.ts

@@ -127,7 +127,7 @@ describe('wire event bridge', () => {
     const bench = await mount()
     let resets = 0
     bench.ctx.on('connection/reset', () => { resets++ })
-    const description = { version: '0', cwd: '/f', attachedSessions: 0, canOpenPath: true }
+    const description = { version: '0', cwd: '/f', attachedSessions: 0, home: '/h', canOpenPath: true }
     bench.sinks?.onConnected?.(description)
     bench.sinks?.onConnected?.(description) // second generation after a reconnect
     expect(resets).toBe(2)

packages/client/ui-conversation/README.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 packages/client/ui-conversation/README.md
-README.md: d9b774bdf5bfc2beaa33fe0d3ada8263b863798d
-README.zh.md: 94da3def811fb901132f53fd6dbf4de0ccd6b3c8
+README.md: 7793273e053142b232cf6e810cd915ddbdeb90b0
+README.zh.md: f45dfa6d72ecf5b49a12c9550d41dbcbc35f9be0

packages/client/ui-conversation/README.md

@@ -22,7 +22,7 @@ Logged non-user messages render as a default-collapsed disclosure whose header n
 
 A Think row stays collapsed by default and exposes live reasoning throughput without expanding the chain of thought: while its reasoning block is the streaming tail, the summary switches from the settled first line to the latest non-blank line and its one-line scrollport follows each delta to the inline end. Expanding the row removes the moving summary and leaves the full reasoning in ordinary page flow, so page reading never fights an internal follower; settlement restores the stable first-line summary at the left edge ([decision](../../../.agents/notes/implemented/feature/2026-08-02-web-thinking-tail-scroll.md)).
 
-The chat view keeps Tool placement but delegates Tool presentation. Each ordered `tool-call` Conversation Node dispatches through the matching key of `conversation.chat.node`, while the details shell passes the selected call through `conversation.details.tool`. The assembled Web bundle registers [`ui-tool`](../ui-tool/README.md) for that Chat Node key; it renders the Runtime-projected recursive root/child tree and owns per-name dispatch, generic rendering, and render-intent cards. The details seat alone retains a raw-result fallback when that renderer is absent.
+The chat view keeps Tool placement but delegates Tool presentation. Each ordered `tool-call` Conversation Node dispatches through the matching key of `conversation.chat.node`, while the details shell passes the selected call through `conversation.details.tool`. The assembled Web bundle registers [`ui-tool`](../ui-tool/README.md) for that Chat Node key; it renders the Runtime-projected recursive root/child tree and owns per-name dispatch, generic rendering, and render-intent cards. The details seat alone retains a raw-result fallback when that renderer is absent. A path click through the injected `openFile` asks the Host to open that path (relative paths resolve against the session cwd). A Host or OS refusal opens an in-page dialog with the thrown reason and a Retry of the same path; Cancel, Escape, the close control, and a mask click dismiss it ([decision](../../../.agents/notes/implemented/bug-fix/2026-08-18-tool-row-file-open-failure.md)).
 
 The chat flow projects consecutive model-retry nodes across retry turns into one stable, muted status row updated to the latest attempt; every retry event remains in the runtime snapshot and session log. Its frontend countdown anchors the scheduled delay to client receipt, avoiding host/browser clock skew, rounds remaining time up to seconds, and has a one-second floor. The latest unresolved retry uses a left-to-right text shimmer. Subsequent turn facts distinguish an attempt that started from one cancelled during backoff, while the Host running bit only controls the live animation; the row then shows a static completed or cancelled label. Normal policy rows show the finite retry maximum; always policy rows show `∞`. Activating the row reveals the latest exact retry delay and failure message. The client runtime removes each failed step's streaming tail before its retry node arrives, while the status remains visible after a later attempt succeeds. An unretried terminal failure renders as a persistent inline status at its turn boundary, showing the display-safe durable message and optional error code without offering an action the Host cannot fulfill; AUTH copy never echoes provider-supplied credential fragments.
 

packages/client/ui-conversation/README.zh.md

@@ -20,7 +20,7 @@ Chat 业务行是彼此独立的注册表贡献，不是封闭的内建联合。
 
 Think 行默认保持折叠，并在不展开思维链的情况下暴露实时推理（reasoning）吞吐：当推理块是流式输出尾部时，摘要从结算后的首行切换到最新的非空行，其单行滚动区会随每个 delta 追到行内末端。展开该行会移除移动摘要，让完整推理进入普通页面流，因此页面阅读不会与内部跟随器争夺滚动；结算后恢复左对齐的稳定首行摘要（[决策](../../../.agents/notes/implemented/feature/2026-08-02-web-thinking-tail-scroll.md)）。
 
-聊天视图保留工具的消息流位置，但委托其展示。每个已排序的 `tool-call` Conversation Node 都通过 `conversation.chat.node` 的同名 key 分发；详情壳层则通过 `conversation.details.tool` 传递当前选中的调用。组装后的 Web bundle 为该 Chat Node key 注册 [`ui-tool`](../ui-tool/README.md)，由后者渲染运行时已投影的递归 root/child 树，并负责按名称分发、通用展示和 render-intent 卡片；只有详情席位会在该 renderer 缺席时保留 raw-result fallback。
+聊天视图保留工具的消息流位置，但委托其展示。每个已排序的 `tool-call` Conversation Node 都通过 `conversation.chat.node` 的同名 key 分发；详情壳层则通过 `conversation.details.tool` 传递当前选中的调用。组装后的 Web bundle 为该 Chat Node key 注册 [`ui-tool`](../ui-tool/README.md)，由后者渲染运行时已投影的递归 root/child 树，并负责按名称分发、通用展示和 render-intent 卡片；只有详情席位会在该 renderer 缺席时保留 raw-result fallback。经注入的 `openFile` 点击路径会请 Host 打开该路径（相对路径按会话 cwd 解析）。Host 或操作系统拒绝时，页面内对话框展示抛出的原因，并提供对同一路径的重试；取消、Escape、关闭控件和点击遮罩会关掉对话框（[决策](../../../.agents/notes/implemented/bug-fix/2026-08-18-tool-row-file-open-failure.md)）。
 
 聊天流会将跨重试轮次连续出现的模型重试节点投影为一个稳定的弱化状态行，并用最新一次尝试更新该行；每个重试事件仍保留在运行时快照与会话日志中。前端倒计时以客户端收到事件的时刻为计划延迟的起点，避免 Host 与浏览器的时钟偏差；剩余时间向上取整到秒，且下限为 1 秒。最近一次尚未完成的重试会显示从左到右的文字渐变动画。后续轮次事实用于区分已开始的尝试与在退避期间取消的尝试，Host 的 running 位只控制实时动画；随后该行会显示静态的已完成或已取消标签。normal 策略行显示有限重试上限；always 策略行显示 `∞`。激活该行会显示最近一次重试的精确延迟和失败消息。客户端运行时会在相应重试节点到达前移除每个失败步骤的流式输出尾部；后续某次尝试成功后，该状态仍保持可见。未进入重试的终态失败会在其轮次边界渲染为持久的内联状态，展示适合显示的持久消息与可选错误码，但不会提供 Host 无法兑现的操作；AUTH 文案绝不会回显提供方给出的凭据片段。
 

packages/client/ui-conversation/src/client/apply.ts

@@ -396,10 +396,7 @@ export function apply(ctx: Context): void {
         fileMentions: owner => ctx.get('chatFileMentions')?.forClosing(owner),
         openFile: (path) => {
           const cwd = sessions.list.getSnapshot().byId[sessionId]?.cwd
-          void workspaces.openPath(resolveWorkspacePath(cwd, path)).catch(() => {
-            // Host/OS open failures stay silent in the chat row; the native
-            // app surfaces its own error dialog when the path is unusable.
-          })
+          return workspaces.openPath(resolveWorkspacePath(cwd, path))
         },
         loadOlder: () => { void scoped.loadOlder() },
         loadImage: attachment => conversation.resolveImage(sessionId, attachment),

packages/client/ui-conversation/src/client/chat/ChatView.module.css

@@ -190,3 +190,8 @@
 .toBottom:hover {
   background: var(--dsw-alias-button-floating-hover);
 }
+
+/* Host open-path refusal: same dialog family as the workspace folder error. */
+.modalAction {
+  min-width: 72px;
+}

packages/client/ui-conversation/src/client/chat/ChatView.tsx

@@ -1,7 +1,8 @@
 // ChatView: the default conversation view — one stable keyed parent list over
 // final business Nodes, plus paging, pending steering and bottom-follow.
 // Each row dispatches through 'conversation.chat.node'; ui-tool owns the
-// tool-call renderer and its recursive root/subcall composition.
+// tool-call renderer and its recursive root/subcall composition. A Host
+// open-path refusal from the injected opener is an in-page dialog here.
 //
 // Scroll: when nested under `[data-conversation-scroll]` (active conversation
 // column), that host is the scrollport and this view is flow content; when
@@ -14,7 +15,7 @@
 
 import { useCallback, useEffect, useLayoutEffect, useMemo, useRef, useState } from 'react'
 import type { ConversationTimelineSnapshot } from '@deepseek-ai/dsh-client-runtime/client'
-import { IconChevronDownOutline14 } from '@deepseek-ai/dsh-client-ui-primitives'
+import { Button, IconChevronDownOutline14, Modal } from '@deepseek-ai/dsh-client-ui-primitives'
 import type { ChatViewSlotProps, RenderMessageImages } from '../contract/slots.ts'
 import { PendingSteeringBubble } from './MessageItem.tsx'
 import { ChatNodeSeat } from './ChatNodeSeat.tsx'
@@ -95,6 +96,17 @@ function scrollPosition(list: HTMLElement, scrollport: HTMLElement): ChatScrollP
   }
 }
 
+/** Host/OS refusal text for the file-open dialog; empty throws keep a locale fallback. */
+function openFailureMessage(error: unknown, fallback: string): string {
+  const message = error instanceof Error ? error.message : String(error)
+  return message === '' ? fallback : message
+}
+
+/** ProducedFiles opens the session workspace as `.`. */
+function isFolderOpenPath(path: string): boolean {
+  return path === '.'
+}
+
 function runningTurnStartTime(timeline: ConversationTimelineSnapshot): number | null {
   let latest: number | null = null
   for (const turn of timeline.turns.values()) {
@@ -159,6 +171,40 @@ export function ChatView({
   const hasMore = useSession(s => s.hasMore)
   const loadingOlder = useSession(s => s.loadingOlder)
   const selectedCallId = useStore(s => s.selection?.callId)
+  const [fileOpenError, setFileOpenError] = useState<{ path: string; message: string } | null>(null)
+  const [fileOpenBusy, setFileOpenBusy] = useState(false)
+  // Close/retry must ignore a settlement that started before the latest
+  // gesture; otherwise a cancelled in-flight refusal reopens the dialog.
+  const fileOpenRequest = useRef(0)
+
+  const requestOpenFile = useCallback((path: string) => {
+    const id = ++fileOpenRequest.current
+    setFileOpenBusy(true)
+    void openFile(path).then(
+      () => {
+        if (id !== fileOpenRequest.current) return
+        setFileOpenError(null)
+        setFileOpenBusy(false)
+      },
+      (error: unknown) => {
+        if (id !== fileOpenRequest.current) return
+        setFileOpenError({
+          path,
+          message: openFailureMessage(
+            error,
+            t(isFolderOpenPath(path) ? 'fileOpen.folderUnknown' : 'fileOpen.unknown'),
+          ),
+        })
+        setFileOpenBusy(false)
+      },
+    )
+  }, [openFile, t])
+
+  const closeFileOpenError = useCallback(() => {
+    fileOpenRequest.current += 1
+    setFileOpenError(null)
+    setFileOpenBusy(false)
+  }, [])
 
   const pendingSteering = useMemo(
     () => inbox.filter(item => item.placement === 'steering'),
@@ -390,7 +436,7 @@ export function ChatView({
               useSession={useSession}
               selectedCallId={selectedCallId}
               cwd={cwd}
-              openFile={openFile}
+              openFile={requestOpenFile}
               inspectCall={inspectCall}
               forkAt={forkAt}
               renderMessageImages={renderMessageImages}
@@ -431,6 +477,44 @@ export function ChatView({
           </div>
         )}
       </div>
+      {fileOpenError !== null && (
+        <FileOpenErrorDialog
+          path={fileOpenError.path}
+          message={fileOpenError.message}
+          busy={fileOpenBusy}
+          onClose={closeFileOpenError}
+          onRetry={() => { requestOpenFile(fileOpenError.path) }}
+          t={t}
+        />
+      )}
     </div>
   )
 }
+
+/** In-page Host open-path refusal: the wire reason plus a retry of the same path. */
+function FileOpenErrorDialog({
+  path, message, busy, onClose, onRetry, t,
+}: {
+  path: string
+  message: string
+  busy: boolean
+  onClose: () => void
+  onRetry: () => void
+  t: ChatViewSlotProps['t']
+}) {
+  return (
+    <Modal
+      open
+      onClose={onClose}
+      closeLabel={t('close')}
+      title={t(isFolderOpenPath(path) ? 'fileOpen.folderTitle' : 'fileOpen.title')}
+      description={message}
+      footer={(
+        <>
+          <Button variant="outline" className={css.modalAction} onClick={onClose}>{t('cancel')}</Button>
+          <Button variant="primary" className={css.modalAction} disabled={busy} onClick={onRetry}>{t('retry')}</Button>
+        </>
+      )}
+    />
+  )
+}

packages/client/ui-conversation/src/client/contract/slots.ts

@@ -714,9 +714,11 @@ export interface ChatViewInjected {
   openDetails: (target: SelectionTarget) => void
   /**
    * Open a tool-arg filesystem path with the host OS default application
-   * (relative paths resolve against the session cwd).
+   * (relative paths resolve against the session cwd). Always returns a
+   * promise: fulfills when the Host opens the path, rejects when it cannot
+   * hand the path off (the chat view shows that reason and a retry).
    */
-  openFile: (path: string) => void
+  openFile: (path: string) => Promise<void>
   loadOlder: () => void
   /** Resolve a session-authorized historical image for inline display. */
   loadImage: (attachment: ImageAttachmentRef) => Promise<string>

packages/client/ui-conversation/src/client/locales.ts

@@ -92,6 +92,10 @@ export const zh = {
   'chat.loadError': '历史加载失败：{message}（{code}）',
   'chat.loadOlder': '加载更早',
   'chat.toBottom': '回到底部',
+  'fileOpen.title': '无法打开文件',
+  'fileOpen.unknown': '无法打开此文件',
+  'fileOpen.folderTitle': '无法打开文件夹',
+  'fileOpen.folderUnknown': '无法打开此文件夹',
   'message.extraBlock': '附加内容块',
   'message.contextInjection': '上下文注入',
   'message.contextRecall': '跨会话召回',
@@ -263,6 +267,10 @@ export const en = {
   'chat.loadError': 'Failed to load history: {message} ({code})',
   'chat.loadOlder': 'Load earlier',
   'chat.toBottom': 'Back to bottom',
+  'fileOpen.title': 'Couldn’t open file',
+  'fileOpen.unknown': 'Couldn’t open this file',
+  'fileOpen.folderTitle': 'Couldn’t open folder',
+  'fileOpen.folderUnknown': 'Couldn’t open this folder',
   'message.extraBlock': 'Extra content block',
   'message.contextInjection': 'Context injection',
   'message.contextRecall': 'Session recall',

packages/client/ui-conversation/tests/apply-inject.client.spec.tsx

@@ -233,13 +233,21 @@ describe('conversation slot inject API', () => {
   it('openFile (chat view face) resolves against session cwd and calls workspaces.openPath', async () => {
     const b = await bench()
     const { injected } = b.chatViewApi(ROOT)
-    injected.openFile('src/a.ts')
+    await injected.openFile('src/a.ts')
     await vi.waitFor(() => {
       expect(b.runtime.workspaces.calls).toContainEqual({ method: 'openPath', args: ['/proj/src/a.ts'] })
     })
     await b.runtime.dispose()
   })
 
+  it('openFile rejects when the Host cannot open the path', async () => {
+    const b = await bench()
+    b.runtime.workspaces.stub('openPath', () => Promise.reject(new Error('xdg-open is not available')))
+    const { injected } = b.chatViewApi(ROOT)
+    await expect(injected.openFile('src/a.ts')).rejects.toThrow('xdg-open is not available')
+    await b.runtime.dispose()
+  })
+
   it('routes workspace switching through the runtime owner, carrying the draft', async () => {
     const b = await bench()
     const resident = b.residentApi(ROOT)

packages/client/ui-conversation/tests/chat-view.client.spec.tsx

@@ -4,7 +4,7 @@
 // ObservableSnapshot fake, no wire or Tool presentation plugin.
 
 import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest'
-import { act, cleanup, fireEvent, render, within } from '@testing-library/react'
+import { act, cleanup, fireEvent, render, screen, waitFor, within } from '@testing-library/react'
 import { useEffect } from 'react'
 import type {
   AssistantMessageNode, CommandNode, CompactionSummaryNode, ConversationNode, ConversationSnapshot,
@@ -152,7 +152,7 @@ function emptyWorkspaces() {
 function makeHarness(init?: Partial<ConversationSnapshot>) {
   const { set, source } = makeSource(init)
   const openDetails = vi.fn<(t: SelectionTarget) => void>()
-  const openFile = vi.fn<(path: string) => void>()
+  const openFile = vi.fn<(path: string) => Promise<void>>().mockResolvedValue(undefined)
   const loadOlder = vi.fn()
   const inspectCall = vi.fn<(callId: string) => void>()
   // In-memory scroll memory matching the apply.ts per-session map contract.
@@ -348,7 +348,7 @@ describe('Chat node rendering', () => {
       resolve: (value) => {
         if (value !== 'report.html') return undefined
         return {
-          open: () => { h.openFile(`for-seq-${String(owner.seq)}/site/report.html`) },
+          open: () => { void h.openFile(`for-seq-${String(owner.seq)}/site/report.html`) },
           label: '打开 site/report.html',
           title: 'site/report.html',
         }
@@ -967,10 +967,113 @@ describe('ChatView', () => {
     })
     const owner = calls[0]?.owner as RoutedChatNodeOwner
     expect((owner.node.data as { readonly root: ToolCallBlock }).root).toBe(block)
-    expect(owner.openFile).toBe(h.openFile)
+    expect(owner.openFile).not.toBe(h.openFile)
+    owner.openFile('src/a.ts')
+    expect(h.openFile).toHaveBeenCalledWith('src/a.ts')
     expect(owner.inspectCall).toBe(h.inspectCall)
   })
 
+  it('shows a Host open refusal with the reason and retries the same path', async () => {
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce(new Error('xdg-open is not available'))
+      .mockResolvedValueOnce(undefined)
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('src/a.ts') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件' })).toBeTruthy()
+    })
+    expect(screen.getByRole('dialog', { name: '无法打开文件' }).textContent).toContain('xdg-open is not available')
+    await act(async () => { fireEvent.click(screen.getByRole('button', { name: '重试' })) })
+    await waitFor(() => {
+      expect(screen.queryByRole('dialog')).toBeNull()
+    })
+    expect(openFile).toHaveBeenCalledTimes(2)
+    expect(openFile).toHaveBeenNthCalledWith(1, 'src/a.ts')
+    expect(openFile).toHaveBeenNthCalledWith(2, 'src/a.ts')
+  })
+
+  it('keeps a non-Error Host refusal visible and dismisses it on cancel', async () => {
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce('permission denied')
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('notes.md') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件' }).textContent).toContain('permission denied')
+    })
+    fireEvent.click(screen.getByRole('button', { name: '取消' }))
+    expect(screen.queryByRole('dialog')).toBeNull()
+    expect(openFile).toHaveBeenCalledTimes(1)
+  })
+
+  it('substitutes the unknown-open copy when the Host refusal has no text', async () => {
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce(new Error(''))
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('empty.ts') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件' }).textContent).toContain('无法打开此文件')
+    })
+  })
+
+  it('names a workspace-folder Host refusal as a folder', async () => {
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce(new Error(''))
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('.') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件夹' }).textContent).toContain('无法打开此文件夹')
+    })
+  })
+
+  it('ignores a Host refusal that settles after the dialog is dismissed', async () => {
+    let rejectRetry!: (error: unknown) => void
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce(new Error('first refusal'))
+      .mockImplementationOnce(() => new Promise<void>((_resolve, reject) => {
+        rejectRetry = reject
+      }))
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('src/a.ts') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件' }).textContent).toContain('first refusal')
+    })
+    await act(async () => { fireEvent.click(screen.getByRole('button', { name: '重试' })) })
+    fireEvent.click(screen.getByRole('button', { name: '取消' }))
+    expect(screen.queryByRole('dialog')).toBeNull()
+    await act(async () => { rejectRetry(new Error('late refusal')) })
+    expect(screen.queryByRole('dialog')).toBeNull()
+  })
+
+  it('ignores a Host open that succeeds after the dialog is dismissed', async () => {
+    let resolveRetry!: () => void
+    const openFile = vi.fn<(path: string) => Promise<void>>()
+      .mockRejectedValueOnce(new Error('first refusal'))
+      .mockImplementationOnce(() => new Promise<void>((resolve) => {
+        resolveRetry = () => { resolve() }
+      }))
+    const h = makeHarness({ nodes: [toolResult(3, 'a')] })
+    h.props.openFile = openFile
+    render(<h.ChatView {...h.props} />)
+    await act(async () => { h.toolOwners[0]!.openFile('src/a.ts') })
+    await waitFor(() => {
+      expect(screen.getByRole('dialog', { name: '无法打开文件' }).textContent).toContain('first refusal')
+    })
+    await act(async () => { fireEvent.click(screen.getByRole('button', { name: '重试' })) })
+    fireEvent.click(screen.getByRole('button', { name: '取消' }))
+    await act(async () => { resolveRetry() })
+    expect(screen.queryByRole('dialog')).toBeNull()
+  })
+
   it('prepend preserves a semantic row; a trailing user node force-scrolls', () => {
     const h = makeHarness({ nodes: [user(5, 'later'), assistant(6, 'a')], hasMore: true })
     const view = render(<h.ChatView {...h.props} />)

packages/client/ui-conversation/tests/views-type-chain.client.spec.tsx

@@ -45,7 +45,7 @@ describe('view-ring type negatives (compile-time; body never runs)', () => {
         // @ts-expect-error openDetails takes a SelectionTarget, not a string
         props.openDetails('nope')
         // @ts-expect-error openFile takes a path string, not a SelectionTarget
-        props.openFile({ turnSeq: 1, callId: 'c' })
+        void props.openFile({ turnSeq: 1, callId: 'c' })
         return null
       }
       void chatProps

packages/client/ui-deliverables/tests/produced-files.client.spec.tsx

@@ -286,7 +286,7 @@ describe('ProducedFiles row', () => {
   ): Pick<ProducedFilesProps, 'isLoopback' | 'useHostDescription'> => {
     const description = canOpenPath === undefined
       ? undefined
-      : { version: 'test', cwd: '/workspace', attachedSessions: 1, canOpenPath }
+      : { version: 'test', cwd: '/workspace', attachedSessions: 1, home: '/h', canOpenPath }
     return {
       isLoopback,
       useHostDescription: selector => selector(description),

packages/client/ui-message-feedback/README.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 packages/client/ui-message-feedback/README.md
-README.md: 461e87589567eb95b075839d5893cc551e67a035
-README.zh.md: 31f722021ff65f476a6ba1ab0211fd4e091671d2
+README.md: d3bb4e28b95da2fde0d26b7ef83ebca377b4939f
+README.zh.md: d823bb9de2b7d39d0bc5d00164b8ce8dffec4a34

packages/client/ui-message-feedback/README.md

@@ -2,7 +2,7 @@
 
 English | [中文](README.zh.md)
 
-Per-message feedback plugin, browser half: a Like/Dislike pair plus an optional note, contributed as the `feedback` entry (order 10) of the `conversation.chat.assistant-actions` strip. The strip is declared by `ui-conversation` and rendered inside the finalized assistant message's IconActions row, between copy and branch, so the controls inherit that row's chrome and hover behavior. Only finalized messages reach the slot — an interruption-frozen partial carries no `messageId` and therefore no feedback controls. The strip renders once per turn, on the closing assistant message that owns the turn's IconActions row: earlier steps of a multi-step turn produce tool rows rather than a rateable body, so they present no controls even though the Host would accept them as targets.
+Per-message feedback plugin, browser half: a Like/Dislike pair plus an optional note, contributed as the `feedback` entry (order 10) of the `conversation.chat.assistant-actions` strip. The strip is declared by `ui-conversation` and rendered inside the finalized assistant message's IconActions row, between copy and branch, so the controls inherit that row's chrome and hover behavior. The note editor itself does not sit in that row: it is a `role="dialog"` popover portaled to `document.body` and anchored under its trigger, so the row keeps its single line whether the editor is open or closed and the panel is not clipped by the conversation column. A rating or list-load failure shows inline in the row; a note-save failure shows inside the popover, which stays open so the draft can be corrected. Only finalized messages reach the slot — an interruption-frozen partial carries no `messageId` and therefore no feedback controls. The strip renders once per turn, on the closing assistant message that owns the turn's IconActions row: earlier steps of a multi-step turn produce tool rows rather than a rateable body, so they present no controls even though the Host would accept them as targets.
 
 One `MessageFeedbackController` per Session backs every message control in that Session, so a single `messageFeedback.list` read seeds the whole transcript. The read is deferred to the first hover or focus rather than fired on mount, because the controls mount once per settled message in the visible history.
 

packages/client/ui-message-feedback/README.zh.md

@@ -2,7 +2,7 @@
 
 [English](README.md) | 中文
 
-单条消息反馈插件的浏览器侧：一对 Like/Dislike 按钮加一个可选备注，作为 `conversation.chat.assistant-actions` 条带的 `feedback` 条目（order 10）贡献。该条带由 `ui-conversation` 声明，渲染在已定稿助手消息的 IconActions 行内、复制与分支之间，因此控件沿用该行的样式与 hover 行为。只有已定稿的消息能到达这个 slot——被中断冻结的部分输出不带 `messageId`，因此也没有反馈控件。该操作栏每个 Turn 渲染一次，位于持有该 Turn IconActions 行的收尾助手消息上：多步骤 Turn 中较早的步骤产出的是工具行而非可评分正文，因此即使 Host 会接受它们作为目标，界面上也不出现控件。
+单条消息反馈插件的浏览器侧：一对 Like/Dislike 按钮加一个可选备注，作为 `conversation.chat.assistant-actions` 条带的 `feedback` 条目（order 10）贡献。该条带由 `ui-conversation` 声明，渲染在已定稿助手消息的 IconActions 行内、复制与分支之间，因此控件沿用该行的样式与 hover 行为。备注编辑器本身不在这一行里：它是一个 `role="dialog"` 的浮层，portal 到 `document.body` 并锚定在其触发按钮下方，因此无论编辑器是否打开该行都保持单行，面板也不会被会话列裁掉。评分或列表加载失败在行内展示；备注保存失败在浮层内展示，且面板保持打开以便修正草稿。只有已定稿的消息能到达这个 slot——被中断冻结的部分输出不带 `messageId`，因此也没有反馈控件。该操作栏每个 Turn 渲染一次，位于持有该 Turn IconActions 行的收尾助手消息上：多步骤 Turn 中较早的步骤产出的是工具行而非可评分正文，因此即使 Host 会接受它们作为目标，界面上也不出现控件。
 
 每个 Session 一个 `MessageFeedbackController`，支撑该 Session 内所有消息的控件，因此一次 `messageFeedback.list` 读取即可填充整段对话。该读取延迟到首次 hover 或 focus 才发起，而不是在挂载时触发，因为可见历史中每条已结束的消息都会挂载一次控件。
 

packages/client/ui-message-feedback/package.json

@@ -71,6 +71,7 @@
     "@deepseek-ai/cordis": "workspace:^",
     "@testing-library/react": "^16.1.0",
     "@types/react": "~18.3.1",
+    "@types/react-dom": "~18.3.0",
     "react": "^18.2.0",
     "react-dom": "^18.2.0"
   },

packages/client/ui-message-feedback/src/client/MessageFeedbackActions.module.css

@@ -1,6 +1,9 @@
 /* Per-message feedback controls. The rating buttons mirror the shared message
-   IconActions chrome so the strip reads as one row; the note editor is an
-   inline expansion anchored to the same row. */
+   IconActions chrome so the strip reads as one row. The note editor is a
+   popover portaled to document.body and fixed from the note trigger's rect,
+   so it neither competes with the row for inline width nor gets cropped by the
+   conversation column's overflow clip. Surface recipe follows the Menu card:
+   r12, inverted hairline border, shadow-lv3. */
 
 .action {
   display: inline-flex;
@@ -47,29 +50,58 @@
   cursor: pointer;
 }
 
-.noteOpen:hover {
+.noteOpen:hover,
+.noteOpen[aria-expanded='true'] {
   background: var(--dsw-alias-interactive-bg-hover);
   color: var(--dsw-alias-label-secondary);
 }
 
-.noteEditor {
-  display: inline-flex;
-  align-items: flex-start;
-  gap: 6px;
+/* Portal surface: fixed in the viewport, left/top supplied inline from the
+   trigger rect. Portaled panels must layer above modal overlays (z 1000). */
+.notePanel {
+  position: fixed;
+  z-index: 1100;
+  box-sizing: border-box;
+  width: 320px;
+  max-width: min(360px, calc(100vw - 24px));
+  /* The width bound's counterpart. `resize: vertical` on the textarea lets the
+     panel be dragged taller, and a panel taller than the viewport would push
+     the placement clamp's upper bound below its own margin, so `top` would go
+     negative and cut off the panel's head. Both bounds keep the 12px margin
+     the clamp uses. */
+  max-height: calc(100vh - 24px);
+  overflow-y: auto;
+  padding: 8px;
+  display: flex;
+  flex-direction: column;
+  gap: 8px;
+  border: 1px solid var(--dsw-alias-border-inverted);
+  border-radius: 12px;
+  background: var(--dsw-specific-menu);
+  box-shadow: var(--dsw-shadow-lv3);
+  --dsh-scrollbar-thumb: var(--dsw-alias-scrollbar-bg-l2);
+  --dsh-scrollbar-thumb-hover: var(--dsw-alias-scrollbar-hover-l2);
 }
 
 .noteInput {
-  width: 260px;
+  width: 100%;
+  box-sizing: border-box;
   padding: 6px 8px;
-  border: 1px solid var(--dsw-alias-border-secondary);
+  border: 1px solid var(--dsw-alias-border-l2);
   border-radius: 8px;
-  background: var(--dsw-alias-bg-primary);
+  background: var(--dsw-alias-bg-layer-1);
   color: var(--dsw-alias-label-primary);
   font: inherit;
   font-size: 13px;
   resize: vertical;
 }
 
+.noteActions {
+  display: flex;
+  justify-content: flex-end;
+  gap: 6px;
+}
+
 .noteSave,
 .noteCancel {
   height: 28px;
@@ -81,8 +113,12 @@
 }
 
 .noteSave {
-  background: var(--dsw-alias-interactive-bg-primary);
-  color: var(--dsw-alias-label-inverse);
+  background: var(--dsw-alias-button-primary-fill);
+  color: var(--dsw-alias-label-primary-foreground);
+}
+
+.noteSave:hover:not(:disabled) {
+  background: var(--dsw-alias-button-primary-hover);
 }
 
 .noteSave:disabled {
@@ -104,5 +140,5 @@
   padding-left: 4px;
   color: var(--dsw-alias-label-tertiary);
   font-size: 13px;
-  line-height: 28px;
+  line-height: 20px;
 }

packages/client/ui-message-feedback/src/client/MessageFeedbackActions.tsx

@@ -1,23 +1,47 @@
 /**
  * Per-message feedback controls: a Like/Dislike pair plus an optional note.
- * Rendered inside the assistant message's IconActions row, so the buttons
- * reuse that row's chrome and sit between copy and branch.
+ * The buttons render inside the assistant message's IconActions row, so they
+ * reuse that row's chrome and sit between copy and branch. The note editor is
+ * a popover (portaled to `document.body`) anchored to the note trigger, not an
+ * inline expansion: a 260px textarea plus buttons cannot fit the row at any
+ * viewport, and an inline element pushed the branch action and clock out of the
+ * conversation column. Portaling out of the column also escapes its `overflow`
+ * clip, so the panel cannot be cropped or detached from the message it annotates.
  * @module @deepseek-ai/dsh-client-ui-message-feedback/client/MessageFeedbackActions
  */
 
-import { useCallback, useEffect, useRef, useState } from 'react'
 import {
-  IconDislikeOutline16, IconLikeOutline16, Tooltip,
+  useCallback, useEffect, useRef, useState,
+  type CSSProperties,
+} from 'react'
+import { createPortal } from 'react-dom'
+import {
+  IconDislikeOutline16, IconLikeOutline16, Tooltip, useAnchoredPosition,
 } from '@deepseek-ai/dsh-client-ui-primitives'
 import type { MessageFeedbackRating } from '@deepseek-ai/dsh-message-feedback/types'
 import type { MessageFeedbackActionProps } from './slots.ts'
 import css from './MessageFeedbackActions.module.css'
 
+/** Safe distance kept between the panel and the viewport edges (the Menu portal margin). */
+const PANEL_MARGIN = 12
+
+/** Distance between the trigger's bottom edge and the panel's top. */
+const PANEL_GAP = 4
+
+/**
+ * Unplaced portal panel: hidden but laid out so `offsetWidth` is real for the
+ * clamp. The explicit insets match `Menu`'s measure style — a `position: fixed`
+ * element with auto insets otherwise sits at its static position, a different
+ * origin than the one the first placement measures from.
+ */
+const MEASURE_STYLE: CSSProperties = { visibility: 'hidden', left: 0, top: 0 }
+
 /**
  * One message's feedback controls.
  * @param props - the owner's message identity, the injected verbs, and the
  * shared feedback hook.
- * @returns the rating buttons, plus the note editor while it is open.
+ * @returns the rating buttons and the note trigger, with the note editor
+ * portal-open beneath the trigger while it is open.
  */
 export function MessageFeedbackActions({ messageId, ensure, rate, toggle, clearNote, useFeedback, t }: MessageFeedbackActionProps) {
   const item = useFeedback(view => view.items.get(messageId))
@@ -26,7 +50,15 @@ export function MessageFeedbackActions({ messageId, ensure, rate, toggle, clearN
   const [noteOpen, setNoteOpen] = useState(false)
   const [draft, setDraft] = useState('')
   const [pending, setPending] = useState(false)
-  const [failure, setFailure] = useState<string | null>(null)
+  // A rating or load failure surfaces beside the rating buttons, always legible
+  // whether or not the note popover is open.
+  const [rowFailure, setRowFailure] = useState<string | null>(null)
+  // A note save failure surfaces inside the note popover, where the human is
+  // looking; it stays open so the draft survives to be corrected.
+  const [noteFailure, setNoteFailure] = useState<string | null>(null)
+  const triggerRef = useRef<HTMLButtonElement>(null)
+  const panelRef = useRef<HTMLDivElement>(null)
+  const inputRef = useRef<HTMLTextAreaElement>(null)
   // The controls mount for every settled message in the transcript, so the
   // Session's feedback is read once on first hover/focus rather than on mount.
   const seeded = useRef(false)
@@ -39,47 +71,158 @@ export function MessageFeedbackActions({ messageId, ensure, rate, toggle, clearN
   const alive = useRef(true)
   useEffect(() => () => { alive.current = false }, [])
 
-  const settle = useCallback((result: { ok: boolean; error?: { code: string } }) => {
+  /** Bumped whenever an editing session ends, so a late save can tell it is stale. */
+  const noteGeneration = useRef(0)
+
+  /** Current panel open-state, readable from a stale closure via a ref. */
+  const noteOpenRef = useRef(false)
+  useEffect(() => { noteOpenRef.current = noteOpen }, [noteOpen])
+
+  const errorCopy = useCallback((result: { ok: boolean; error?: { code: string } }) => {
+    return result.error?.code === 'version-conflict' ? t('error.conflict') : t('error.generic')
+  }, [t])
+
+  const settleRating = useCallback((result: { ok: boolean; error?: { code: string } }) => {
     if (!alive.current) return
     setPending(false)
-    if (result.ok) {
-      setFailure(null)
-      return
-    }
-    setFailure(result.error?.code === 'version-conflict' ? t('error.conflict') : t('error.generic'))
-  }, [t])
+    setRowFailure(result.ok ? null : errorCopy(result))
+  }, [errorCopy])
+
+  const closeNote = useCallback(() => {
+    // Ends the editing session, so any save still in flight becomes stale.
+    noteGeneration.current += 1
+    setNoteOpen(false)
+  }, [])
 
   const onRate = useCallback((next: MessageFeedbackRating) => {
     setPending(true)
-    setFailure(null)
+    setRowFailure(null)
     // The controller decides retract-vs-replace from the committed item, so a
     // click that lands before the first list read still toggles the stored
     // value instead of this render's empty view.
-    setNoteOpen(false)
-    void toggle(messageId, next).then(settle)
-  }, [messageId, settle, toggle])
+    closeNote()
+    void toggle(messageId, next).then(settleRating)
+  }, [closeNote, messageId, settleRating, toggle])
 
   // The rating is a parameter because only the note editor's render site can
   // prove one is recorded; that removes an unreachable undefined guard here.
   const onSaveNote = useCallback((current: MessageFeedbackRating) => {
     const trimmed = draft.trim()
     setPending(true)
-    setFailure(null)
+    setNoteFailure(null)
+    // A save belongs to the editing session that started it. Closing and
+    // reopening the panel begins a new one, and a late reply from the old
+    // session must not act on it: a stale success would shut the panel the
+    // human just opened, and a stale failure would describe a draft this
+    // session never sent.
+    const generation = noteGeneration.current
+    // What a session reopened before this save commits would be seeded with.
+    const staleSeed = item?.note ?? ''
     // An emptied editor removes the note explicitly; `rate` alone preserves a
     // stored note, so it cannot express deletion.
     const settled = trimmed.length === 0
       ? clearNote(messageId)
       : rate(messageId, current, trimmed)
     void settled.then((result) => {
-      settle(result)
-      if (result.ok && alive.current) setNoteOpen(false)
+      if (!alive.current) return
+      // `pending` tracks the request in flight, not the editing session, so it
+      // is released either way; all three of like, dislike and Save read
+      // `disabled={pending}`, and holding it would lock the row until remount.
+      // Releasing it unconditionally is safe because those three are the only
+      // mutation entries and each is gated by it, so at most one request is ever
+      // in flight. A future entry that bypasses the gate would have to bind
+      // `pending` to the generation instead of clearing it here.
+      setPending(false)
+      if (result.ok) {
+        // Only the session that is still open may act on a success: closing it
+        // already discarded the draft, and reopening seeded a new one.
+        if (generation === noteGeneration.current) {
+          setNoteFailure(null)
+          setNoteOpen(false)
+          return
+        }
+        // A newer session is open, seeded from the note as it read before this
+        // save committed. Resync it so the editor shows what is stored and the
+        // next save cannot overwrite the text that just landed. An edited draft
+        // is the human's, so it is left alone.
+        setDraft(draftNow => (draftNow === staleSeed ? trimmed : draftNow))
+        return
+      }
+      // A failure from the session still on screen belongs in its panel. One
+      // from an abandoned session is reported only when no new session has
+      // taken over: the row then carries it, so a save that failed after the
+      // human walked away is not silently dropped. Writing it into a reopened
+      // panel instead would label the new draft with the old attempt's error.
+      // `noteOpenRef` — not the `noteOpen` this closure was created from — is
+      // read here, because a close+reopen between the save and resolution
+      // leaves this closure with the panel state from when the save started.
+      if (generation === noteGeneration.current || !noteOpenRef.current) {
+        setNoteFailure(errorCopy(result))
+      }
     })
-  }, [clearNote, draft, messageId, rate, settle])
+  }, [clearNote, draft, errorCopy, item?.note, messageId, noteOpenRef, rate])
 
-  const openNote = useCallback(() => {
+  // The trigger toggles: while closed it opens the popover (seeding the draft
+  // with the recorded note), while open it closes it. Toggling closed via the
+  // trigger also fires the outside/within logic correctly because the trigger
+  // is inside the panel's "inside" region.
+  const toggleNote = useCallback(() => {
+    if (noteOpen) {
+      closeNote()
+      return
+    }
     setDraft(item?.note ?? '')
+    // A note-save failure belongs to the editing session that produced it. The
+    // panel stays open on failure so the draft can be corrected, but once it is
+    // closed and reopened the draft is reseeded from the stored note, so a
+    // carried-over error would describe an attempt the new draft never made.
+    // A failure that arrives after the panel closed is reported in the row, and
+    // clearing it here is what retires that notice when a new session starts.
+    setNoteFailure(null)
     setNoteOpen(true)
-  }, [item?.note])
+  }, [noteOpen, closeNote, item?.note])
+
+  // Place the portaled panel from the trigger rect before paint and keep it
+  // with the trigger on scroll/resize, the same anchoring `Menu` uses for its
+  // portal mode.
+  const pos = useAnchoredPosition({
+    open: noteOpen,
+    anchorRef: triggerRef,
+    panelRef,
+    gap: PANEL_GAP,
+    margin: PANEL_MARGIN,
+  })
+
+  // Focus the input and close on Escape or outside pointer-down while open.
+  useEffect(() => {
+    if (!noteOpen) return
+    inputRef.current?.focus()
+    const onPointerDown = (e: PointerEvent) => {
+      if (!(e.target instanceof Node)) return
+      if (triggerRef.current?.contains(e.target) === true) return
+      if (panelRef.current?.contains(e.target) === true) return
+      closeNote()
+    }
+    const onKeyDown = (e: KeyboardEvent) => {
+      if (e.key === 'Escape') closeNote()
+    }
+    document.addEventListener('pointerdown', onPointerDown)
+    document.addEventListener('keydown', onKeyDown)
+    return () => {
+      document.removeEventListener('pointerdown', onPointerDown)
+      document.removeEventListener('keydown', onKeyDown)
+    }
+  }, [noteOpen, closeNote])
+
+  // Return focus to the trigger only when the panel actually closes, not on the
+  // initial mount (a freshly rendered message with a recorded rating must not
+  // pull focus into its action row).
+  const wasOpen = useRef(false)
+  useEffect(() => {
+    if (noteOpen) { wasOpen.current = true; return }
+    if (wasOpen.current) triggerRef.current?.focus()
+    wasOpen.current = false
+  }, [noteOpen])
 
   const likeLabel = rating === 'positive' ? t('action.likeActive') : t('action.like')
   const dislikeLabel = rating === 'negative' ? t('action.dislikeActive') : t('action.dislike')
@@ -116,38 +259,66 @@ export function MessageFeedbackActions({ messageId, ensure, rate, toggle, clearN
           <IconDislikeOutline16 />
         </button>
       </Tooltip>
-      {rating !== undefined && !noteOpen && (
-        <button type="button" className={css.noteOpen} onClick={openNote}>
+      {rating !== undefined && (
+        <button
+          ref={triggerRef}
+          type="button"
+          className={css.noteOpen}
+          aria-haspopup="dialog"
+          aria-expanded={noteOpen}
+          onClick={toggleNote}
+        >
           {item?.note === undefined ? t('note.open') : item.note}
         </button>
       )}
-      {rating !== undefined && noteOpen && (
-        <span className={css.noteEditor}>
+      {rowFailure === null && loadFailed && (
+        <span className={css.failure} role="status">{t('error.load')}</span>
+      )}
+      {rowFailure !== null && <span className={css.failure} role="status">{rowFailure}</span>}
+      {/* A note-save failure normally lives inside the panel, beside the buttons
+          that produced it. Whenever the panel is not on screen it falls back to
+          the row instead: the rating may have disappeared underneath an open
+          editor (another client retracts the feedback, a `version-conflict`
+          reply commits `current: null`, the item goes away), or the human may
+          have closed the panel before a slow save came back. Either way the row
+          reports that the save did not land rather than dropping it. */}
+      {!(rating !== undefined && noteOpen) && noteFailure !== null && (
+        <span className={css.failure} role="status">{noteFailure}</span>
+      )}
+      {rating !== undefined && noteOpen && createPortal(
+        <div
+          ref={panelRef}
+          className={css.notePanel}
+          role="dialog"
+          aria-label={t('note.dialog')}
+          style={pos ?? MEASURE_STYLE}
+        >
           <textarea
+            ref={inputRef}
             className={css.noteInput}
             aria-label={t('note.aria')}
             placeholder={t('note.placeholder')}
             value={draft}
-            rows={2}
+            rows={3}
             onChange={(event) => { setDraft(event.target.value) }}
           />
-          <button
-            type="button"
-            className={css.noteSave}
-            disabled={pending}
-            onClick={() => { onSaveNote(rating) }}
-          >
-            {t('note.save')}
-          </button>
-          <button type="button" className={css.noteCancel} onClick={() => { setNoteOpen(false) }}>
-            {t('note.cancel')}
-          </button>
-        </span>
+          <div className={css.noteActions}>
+            <button
+              type="button"
+              className={css.noteSave}
+              disabled={pending}
+              onClick={() => { onSaveNote(rating) }}
+            >
+              {t('note.save')}
+            </button>
+            <button type="button" className={css.noteCancel} onClick={closeNote}>
+              {t('note.cancel')}
+            </button>
+          </div>
+          {noteFailure !== null && <span className={css.failure} role="status">{noteFailure}</span>}
+        </div>,
+        document.body,
       )}
-      {failure === null && loadFailed && (
-        <span className={css.failure} role="status">{t('error.load')}</span>
-      )}
-      {failure !== null && <span className={css.failure} role="status">{failure}</span>}
     </>
   )
 }

packages/client/ui-message-feedback/src/client/locales.ts

@@ -7,6 +7,7 @@ export const zh = {
   'action.dislike': '有问题的回答',
   'action.dislikeActive': '取消标记',
   'note.open': '补充说明',
+  'note.dialog': '反馈',
   'note.placeholder': '这条回答哪里好，或哪里有问题？（可选）',
   'note.save': '保存',
   'note.cancel': '取消',
@@ -33,6 +34,7 @@ export const en = {
   'action.dislike': 'Bad response',
   'action.dislikeActive': 'Remove rating',
   'note.open': 'Add a note',
+  'note.dialog': 'Feedback',
   'note.placeholder': 'What was good, or what went wrong? (optional)',
   'note.save': 'Save',
   'note.cancel': 'Cancel',

packages/client/ui-message-feedback/tests/message-feedback-actions.client.spec.tsx

@@ -8,7 +8,7 @@
  */
 import { useSyncExternalStore } from 'react'
 import { afterEach, describe, expect, it, vi } from 'vitest'
-import { cleanup, fireEvent, render, waitFor } from '@testing-library/react'
+import { act, cleanup, fireEvent, render, waitFor } from '@testing-library/react'
 import { makeTranslate } from '@deepseek-ai/dsh-client-test-runtime'
 import { zh as commonZh } from '@deepseek-ai/dsh-client-locale/src/locales/zh.ts'
 import type { MessageId } from '@deepseek-ai/dsh-client-connection/client'
@@ -246,4 +246,494 @@ describe('MessageFeedbackActions', () => {
     await waitFor(() => { expect(ui.getByText(zh['error.generic'])).toBeTruthy() })
     expect(ui.queryByText(zh['error.load'])).toBeNull()
   })
+
+  it('portals the note editor to the document body, not into the actions row', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+
+    // The editor must float above the transcript (escaping the conversation
+    // column's overflow clip), so it renders through a portal to document.body
+    // rather than inline inside the component's own container.
+    const panel = ui.getByRole('dialog')
+    expect(panel).toBeTruthy()
+    expect(ui.container.querySelector('[role="dialog"]')).toBeNull()
+    expect(document.body.contains(panel)).toBe(true)
+  })
+
+  it('closes the note popover on Escape', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    fireEvent.keyDown(document, { key: 'Escape' })
+
+    expect(ui.queryByRole('dialog')).toBeNull()
+  })
+
+  it('closes the note popover on an outside pointer-down', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    fireEvent.pointerDown(document.body)
+
+    expect(ui.queryByRole('dialog')).toBeNull()
+  })
+
+  it('keeps the note popover open on a pointer-down inside it', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    const panel = ui.getByRole('dialog')
+    expect(panel).toBeTruthy()
+
+    fireEvent.pointerDown(panel)
+
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('does not close the note popover on a pointer-down on its trigger', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    // The trigger is inside the panel's own region, so pressing it must not be
+    // treated as an outside click; the toggle click below then closes it.
+    fireEvent.pointerDown(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('toggles the note popover closed and open from its trigger', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+    expect(ui.getByLabelText(zh['note.aria'])).toBeTruthy()
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.queryByRole('dialog')).toBeNull()
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('ignores keys other than Escape while the popover is open', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    fireEvent.keyDown(document, { key: 'Enter' })
+
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('publishes no rating-state after the row unmounts mid-flight', async () => {
+    // Directly exercise the early-return of a rating settle once the control has
+    // unmounted: the promise resolution must not touch React state.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: true as const }) }
+    })
+    const view: MessageFeedbackView = { status: 'ready', items: new Map(), error: null }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      toggle: vi.fn(() => gate),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+    const errors: unknown[] = []
+    const onError = (event: ErrorEvent): void => { errors.push(event.error) }
+    window.addEventListener('error', onError)
+
+    fireEvent.click(ui.getByLabelText(zh['action.like']))
+    ui.unmount()
+    release()
+    await gate
+
+    window.removeEventListener('error', onError)
+    expect(errors).toEqual([])
+  })
+
+  it('publishes no note-state after the row unmounts mid-save', async () => {
+    // Same unmount early-return for the note-save settle path: resolving the
+    // save promise after unmount must not touch React state.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: true as const }) }
+    })
+    const view: MessageFeedbackView = { status: 'ready', items: new Map([[MSG, item({ rating: 'positive' })]]), error: null }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const errors: unknown[] = []
+    const onError = (event: ErrorEvent): void => { errors.push(event.error) }
+    window.addEventListener('error', onError)
+
+    const ui = render(<MessageFeedbackActions {...props} />)
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'hi' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+    ui.unmount()
+    release()
+    await gate
+
+    window.removeEventListener('error', onError)
+    expect(errors).toEqual([])
+  })
+
+  it('ignores a pointer-down whose target is not a DOM node', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    // The outside-click guard returns without closing when the event target is
+    // not a DOM node. `document.dispatchEvent` delivers straight to the
+    // document listener, and a non-Node target is not `instanceof Node`.
+    const event = new MouseEvent('pointerdown', { bubbles: true })
+    Object.defineProperty(event, 'target', { configurable: true, value: { notANode: true } })
+    document.dispatchEvent(event)
+
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('returns focus to the trigger when the popover closes', () => {
+    const ui = mount({ current: item({ rating: 'positive' }) })
+    const trigger = ui.getByText(zh['note.open'])
+
+    fireEvent.click(trigger)
+    expect(ui.getByRole('dialog')).toBeTruthy()
+    fireEvent.keyDown(document, { key: 'Escape' })
+
+    // Closing hands focus back, so a keyboard user resumes on the row they
+    // came from rather than at the document root.
+    expect(ui.queryByRole('dialog')).toBeNull()
+    expect(document.activeElement).toBe(trigger)
+  })
+
+  it('does not pull focus when an already-rated message mounts', () => {
+    // The `wasOpen` guard exists for this: a transcript of already-rated
+    // messages must not drag focus into an action row as each one mounts.
+    // Only a real open-then-close returns focus.
+    const elsewhere = document.createElement('button')
+    document.body.append(elsewhere)
+    elsewhere.focus()
+
+    mount({ current: item({ rating: 'positive' }) })
+
+    expect(document.activeElement).toBe(elsewhere)
+    elsewhere.remove()
+  })
+
+  it('drops a stale save failure when the popover is reopened', async () => {
+    // The failure belongs to the editing session that produced it: reopening
+    // reseeds the draft from the stored note, so a carried-over error would
+    // describe an attempt the new draft never made.
+    const ui = mount({
+      current: item({ rating: 'positive' }),
+      rateResult: { ok: false, error: { code: 'note-too-large', message: 'too long' } },
+    })
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'x'.repeat(20) } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+    await waitFor(() => { expect(ui.getByText(zh['error.generic'])).toBeTruthy() })
+
+    fireEvent.keyDown(document, { key: 'Escape' })
+    fireEvent.click(ui.getByText(zh['note.open']))
+
+    expect(ui.queryByText(zh['error.generic'])).toBeNull()
+  })
+
+  it('keeps a save failure visible when the rating disappears underneath it', async () => {
+    // Another client retracts the feedback while the editor is open: the
+    // controller commits `current: null`, the item goes away, and the panel
+    // unmounts. The failure must not vanish with it, so it falls back to the row.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: false, error: { code: 'target-not-found', message: 'gone' } }) }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive' })]]),
+      error: null,
+    }
+    let notify = (): void => {}
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore((cb) => { notify = cb; return () => {} }, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'hi' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    // The retract lands first, then the save rejects.
+    view.items = new Map()
+    notify()
+    release()
+    await gate
+
+    await waitFor(() => { expect(ui.getByText(zh['error.generic'])).toBeTruthy() })
+    expect(ui.queryByRole('dialog')).toBeNull()
+  })
+
+  it('ignores a save that resolves after its editing session ended', async () => {
+    // Closing and reopening starts a new session. A late success from the old
+    // one must not shut the panel the human just opened.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: true as const }) }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive' })]]),
+      error: null,
+    }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'first' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    // Abandon that session and start another before the save lands.
+    fireEvent.keyDown(document, { key: 'Escape' })
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    release()
+    await gate
+    // Flush the `.then` continuation and the render it would cause. Asserted
+    // directly rather than through `waitFor`, which would retry past a panel
+    // that the stale result closed.
+    await act(async () => { await Promise.resolve() })
+
+    expect(ui.getByRole('dialog')).toBeTruthy()
+    // The reply is discarded, but the request is no longer in flight, so the
+    // controls must not stay disabled: `pending` gates the rating buttons and
+    // Save, and leaving it set locks this message's row until it remounts.
+    expect(ui.getByLabelText(zh['action.likeActive']).hasAttribute('disabled')).toBe(false)
+    expect(ui.getByLabelText(zh['action.dislike']).hasAttribute('disabled')).toBe(false)
+    expect(ui.getByText(zh['note.save']).hasAttribute('disabled')).toBe(false)
+  })
+
+  it('reports a save that fails after the human closed the panel', async () => {
+    // A slow save that rejects once the panel is gone must not be swallowed:
+    // the human would otherwise believe the note was stored. With no panel to
+    // show it in, the row carries the notice.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => {
+        resolve({ ok: false, error: { code: 'target-not-found', message: 'gone' } })
+      }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive' })]]),
+      error: null,
+    }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'hi' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    // Walk away before the reply lands, and leave it closed.
+    fireEvent.keyDown(document, { key: 'Escape' })
+    expect(ui.queryByRole('dialog')).toBeNull()
+
+    release()
+    await gate
+    await act(async () => { await Promise.resolve() })
+
+    expect(ui.getByText(zh['error.generic'])).toBeTruthy()
+  })
+
+  it('does not write an abandoned session\'s failure into a reopened panel', async () => {
+    // The old request rejects after the panel was closed and reopened, so the
+    // new session owns the panel. Its draft was not the one that failed, so the
+    // stale error must not be shown there; it belongs to the abandoned session.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => {
+        resolve({ ok: false, error: { code: 'target-not-found', message: 'gone' } })
+      }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive' })]]),
+      error: null,
+    }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText(zh['note.open']))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'first' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    // Abandon that session and start another before the save rejects; unlike
+    // the closed-and-left case, a new panel is now on screen.
+    fireEvent.keyDown(document, { key: 'Escape' })
+    fireEvent.click(ui.getByText(zh['note.open']))
+    expect(ui.getByRole('dialog')).toBeTruthy()
+
+    release()
+    await gate
+    await act(async () => { await Promise.resolve() })
+
+    // The stale failure names a draft the new session never sent, so it stays
+    // out of the reopened panel's status area.
+    expect(ui.queryByText(zh['error.generic'])).toBeNull()
+    expect(ui.getByRole('dialog')).toBeTruthy()
+  })
+
+  it('resyncs an untouched reopened draft to the note that just committed', async () => {
+    // The reopened session seeded from the note as it read before the save
+    // committed, so an untouched draft would show stale text and the next save
+    // could overwrite what just landed.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: true as const }) }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive', note: 'old' })]]),
+      error: null,
+    }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText('old'))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'saved text' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    // Close and reopen before the save lands: the new draft is seeded from the
+    // still-stale stored note.
+    fireEvent.keyDown(document, { key: 'Escape' })
+    fireEvent.click(ui.getByText('old'))
+    expect((ui.getByLabelText(zh['note.aria']) as HTMLTextAreaElement).value).toBe('old')
+
+    release()
+    await gate
+    await act(async () => { await Promise.resolve() })
+
+    expect((ui.getByLabelText(zh['note.aria']) as HTMLTextAreaElement).value).toBe('saved text')
+  })
+
+  it('leaves a reopened draft alone once the human has edited it', async () => {
+    // The opposite arm: an edited draft belongs to the human, so a late save
+    // must not overwrite what they are typing.
+    let release = (): void => {}
+    const gate = new Promise<MessageFeedbackActionResult>((resolve) => {
+      release = () => { resolve({ ok: true as const }) }
+    })
+    const view: MessageFeedbackView = {
+      status: 'ready',
+      items: new Map([[MSG, item({ rating: 'positive', note: 'old' })]]),
+      error: null,
+    }
+    const useFeedback = (<T,>(select: (v: MessageFeedbackView) => T): T =>
+      useSyncExternalStore(() => () => {}, () => select(view))) as never
+    const props = {
+      messageId: MSG,
+      ensure: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      rate: vi.fn(() => gate),
+      toggle: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clearNote: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      clear: vi.fn(() => Promise.resolve<MessageFeedbackActionResult>({ ok: true })),
+      useFeedback,
+      t,
+    } as unknown as Parameters<typeof MessageFeedbackActions>[0]
+    const ui = render(<MessageFeedbackActions {...props} />)
+
+    fireEvent.click(ui.getByText('old'))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'saved text' } })
+    fireEvent.click(ui.getByText(zh['note.save']))
+
+    fireEvent.keyDown(document, { key: 'Escape' })
+    fireEvent.click(ui.getByText('old'))
+    fireEvent.change(ui.getByLabelText(zh['note.aria']), { target: { value: 'my new words' } })
+
+    release()
+    await gate
+    await act(async () => { await Promise.resolve() })
+
+    expect((ui.getByLabelText(zh['note.aria']) as HTMLTextAreaElement).value).toBe('my new words')
+  })
 })

packages/client/ui-message-feedback/tests/styles.client.spec.ts

@@ -0,0 +1,93 @@
+/**
+ * Feedback controls stylesheet contract, asserted against the CSS text on disk.
+ *
+ * A `--dsw-*` name the theme never declares fails silently, and for this sheet
+ * it failed loudly in the product: `border`, `background`, and the primary
+ * button's fill and label each named a token that does not exist, so every one
+ * of those declarations was invalid at computed-value time and dropped. The
+ * note editor shipped with no border and no surface, and its Save button with
+ * neither fill nor readable label. Nothing downstream reports this — the sheet
+ * parses, the classes attach, and the DOM snapshots are unchanged.
+ *
+ * The editor is a popover portaled to `document.body` and fixed-positioned
+ * from the note trigger's rect, so it never enters the IconActions row's flex
+ * layout at all — the row keeps its single 28px line of icons and the note
+ * trigger, and no wrapping (`flex-wrap`) or `order` is needed for it. The
+ * width-independent half of that contract is asserted here (the panel is a
+ * fixed portal, not an inline flex item); the resulting geometry is measured
+ * in a real engine by `apps/web/tests/message-feedback-layout`.
+ */
+import { readdirSync, readFileSync } from 'node:fs'
+import { fileURLToPath } from 'node:url'
+import { describe, expect, it } from 'vitest'
+
+const css = readFileSync(
+  fileURLToPath(new URL('../src/client/MessageFeedbackActions.module.css', import.meta.url)),
+  'utf8',
+)
+// The theme package maps `./styles/*` to `./src/styles/*`, so the declarations
+// stay on the source plane rather than needing a build. Every theme sheet, not
+// just the platform tokens: font and scrollbar variables are declared in
+// siblings, and a gate reading one file would call their names undeclared.
+const tokens = readdirSync(fileURLToPath(new URL('../../ui-theme/src/styles/', import.meta.url)))
+  .filter(name => name.endsWith('.css'))
+  .map(name => readFileSync(fileURLToPath(new URL(`../../ui-theme/src/styles/${name}`, import.meta.url)), 'utf8'))
+  .join('\n')
+
+/**
+ * The declarations of one top-level rule, by selector.
+ * @param selector - the class selector to read, including its leading dot.
+ * @returns the rule's declaration text.
+ */
+function block(selector: string): string {
+  const match = new RegExp(`^\\${selector} \\{([^}]*)\\}`, 'm').exec(css)
+  if (match === null) throw new Error(`MessageFeedbackActions.module.css has no \`${selector}\` rule`)
+  return match[1] ?? ''
+}
+
+describe('MessageFeedbackActions theme styles', () => {
+  it('names only theme variables the token sheet defines', () => {
+    // The regression that motivated this file. An undeclared custom property
+    // has no fallback and does not inherit a usable value: the entire
+    // declaration is thrown away, so the control renders as if the line had
+    // never been written. Every theme-variable prefix the sheets actually use,
+    // not just `--dsw-`: a `--dsh-` name reads as a plausible sibling and would
+    // otherwise slip past into an invalid declaration.
+    const named = [...css.matchAll(/var\((--(?:dsw|dsh|ds)-[a-z0-9-]+)/g)].map(match => match[1])
+    // Vacuity guard: the sheet has to actually name tokens, or the filter below
+    // is satisfied by an empty list and this test proves nothing.
+    expect(named.length).toBeGreaterThan(5)
+    const undeclared = [...new Set(named)].filter(name => !tokens.includes(`  ${String(name)}:`))
+    expect(undeclared).toEqual([])
+  })
+
+  it('never falls back to a literal colour', () => {
+    // A token that resolves is never the problem; an undeclared one takes this
+    // branch, and a literal here is a single colour for both themes.
+    expect(css).not.toMatch(/var\(--dsw-[a-z0-9-]+\s*,\s*(?:#|rgb|rgba|hsl|hsla)/)
+  })
+
+  it('keeps the note editor out of the row as a fixed portal, not a flex item', () => {
+    // The editor is a popover portaled to document.body, so the IconActions row
+    // never has to grow or wrap around it. Fixed positioning comes from the
+    // placement code (inline `left`/`top`), not a class, so only `position:
+    // fixed` and the elevated surface live in the sheet — plus the absence of a
+    // flex rule on the panel, which would resurrect the row-overflow defect an
+    // inline editor had. The row stays one 28px line, so a fixed `width` on the
+    // panel is fine (it floats, it does not compete for row space).
+    expect(block('.notePanel')).toMatch(/position:\s*fixed/)
+    // The panel flex-sets its own children (textarea over buttons), which is
+    // fine. What must be absent is the flex-SIZING that made an inline editor a
+    // row item: grow/shrink/basis (or the `flex:` shorthand) would let it rejoin
+    // the IconActions layout, resurrecting the overflow defect.
+    expect(block('.notePanel')).not.toMatch(/flex-(?:grow|shrink|basis)\s*:/)
+    expect(block('.notePanel')).not.toMatch(/^\s*flex\s*:/m)
+  })
+
+  it('closes every block, so no rule is swallowed by the one above it', () => {
+    // A missing `}` is not a parse error: every rule after it silently becomes
+    // part of the block above, and the controls would paint unstyled.
+    const bare = css.replace(/\/\*[\s\S]*?\*\//g, '')
+    expect((bare.match(/\}/g) ?? []).length).toBe((bare.match(/\{/g) ?? []).length)
+  })
+})

packages/client/ui-primitives/README.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 packages/client/ui-primitives/README.md
-README.md: 9a5b33e6b2ecae0652d640d2a6927e3f1d05b1a1
-README.zh.md: a94935235790504bbeb7f5091e8c8fba86e1c903
+README.md: a675b3cd0aa9e09e243b69065110e1d2b67ff1d9
+README.zh.md: aa67993ec879635fc0677501665d2e23de996dcf

packages/client/ui-primitives/README.md

@@ -2,7 +2,7 @@
 
 English | [中文](README.zh.md)
 
-Pure React atoms (zero cordis): StateDot, DisclosureRow, ic_ds_* icons, Button/Pill/Menu/Modal/Input, the Toast transient banner, the OnboardingSurface first-run takeover (body-portaled mask + opaque stage that holds `#root` inert for exactly its own lifetime), the markdown family (MessageText/MarkdownText/JsonBlock), the read-only JsonTree inspector, the `useAnchoredMaxHeight` hook that clamps a bottom-anchored overlay to the viewport space above its anchor (re-measured on resize, scroll, and a caller-supplied dependency), TerminalBlock, DiffBlock, ReadBlock, SearchBlock, and WebBlock.
+Pure React atoms (zero cordis): StateDot, DisclosureRow, ic_ds_* icons, Button/Pill/Menu/Modal/Input, the Toast transient banner, the OnboardingSurface first-run takeover (body-portaled mask + opaque stage that holds `#root` inert for exactly its own lifetime), the markdown family (MessageText/MarkdownText/JsonBlock), the read-only JsonTree inspector, the `useAnchoredMaxHeight` hook that clamps a bottom-anchored overlay to the viewport space above its anchor (re-measured on resize, scroll, and a caller-supplied dependency), the `useAnchoredPosition` hook that holds a fixed-position floating panel under its anchor (measure, offset, clamp inside the viewport margin, re-placed on capture-phase scroll, window resize, and the panel's own size changes), TerminalBlock, DiffBlock, ReadBlock, SearchBlock, and WebBlock.
 
 ## Hover cards
 

packages/client/ui-primitives/README.zh.md

@@ -2,7 +2,7 @@
 
 [English](README.md) | 中文
 
-纯 React 原子组件（零 cordis）：StateDot、DisclosureRow、ic_ds_* 图标、Button/Pill/Menu/Modal/Input、Toast 短时横幅、OnboardingSurface 首次使用接管层（portal 到 body 的遮罩加不透明展示层，在且仅在自身生命周期内保持 `#root` 为 `inert`）、markdown 家族（MessageText/MarkdownText/JsonBlock）、只读 JsonTree 检查器、`useAnchoredMaxHeight` 钩子（把底部锚定的浮层高度收敛到锚点上方的视口空间，并在 resize、scroll 与调用方提供的依赖变化时重新测量）、TerminalBlock、DiffBlock、ReadBlock、SearchBlock，以及 WebBlock。
+纯 React 原子组件（零 cordis）：StateDot、DisclosureRow、ic_ds_* 图标、Button/Pill/Menu/Modal/Input、Toast 短时横幅、OnboardingSurface 首次使用接管层（portal 到 body 的遮罩加不透明展示层，在且仅在自身生命周期内保持 `#root` 为 `inert`）、markdown 家族（MessageText/MarkdownText/JsonBlock）、只读 JsonTree 检查器、`useAnchoredMaxHeight` 钩子（把底部锚定的浮层高度收敛到锚点上方的视口空间，并在 resize、scroll 与调用方提供的依赖变化时重新测量）、`useAnchoredPosition` 钩子（让固定定位的浮动面板跟住锚点：测量、偏移、按视口边距钳制，并在捕获阶段滚动、窗口缩放与面板自身尺寸变化时重新定位）、TerminalBlock、DiffBlock、ReadBlock、SearchBlock，以及 WebBlock。
 
 ## 悬浮卡片
 

packages/client/ui-primitives/src/index.ts

@@ -13,6 +13,8 @@ export { Input } from './Input.tsx'
 export { Menu } from './Menu.tsx'
 export type { MenuEntry, MenuItem, MenuSeparator, MenuLabel } from './Menu.tsx'
 export { useAnchoredMaxHeight } from './useAnchoredMaxHeight.ts'
+export { useAnchoredPosition } from './useAnchoredPosition.ts'
+export type { AnchoredPositionOptions } from './useAnchoredPosition.ts'
 export { useDismissOnOutsidePointer } from './useDismissOnOutsidePointer.ts'
 export { HoverCard } from './HoverCard.tsx'
 export { Modal } from './Modal.tsx'

packages/client/ui-primitives/src/useAnchoredPosition.ts

@@ -0,0 +1,81 @@
+/**
+ * Keep a fixed-position floating element anchored to a trigger.
+ *
+ * A portaled panel is positioned from its anchor's viewport rect, which stops
+ * being true the moment anything scrolls or the window resizes. This owns that
+ * one concern: measure the anchor, offset the panel below it, clamp the result
+ * inside the viewport, and re-run on scroll (capture phase, so scrollers nested
+ * inside the page are caught too), on resize, and on the panel's own size
+ * changes while the element is open.
+ * @module @deepseek-ai/dsh-client-ui-primitives/useAnchoredPosition
+ */
+
+import { useLayoutEffect, useState, type CSSProperties, type RefObject } from 'react'
+
+/** Inputs for {@link useAnchoredPosition}. */
+export interface AnchoredPositionOptions {
+  /** Whether the floating element is mounted and should track its anchor. */
+  open: boolean
+  /** The element the panel is placed from. */
+  anchorRef: RefObject<HTMLElement | null>
+  /** The floating element, measured so the clamp uses real dimensions. */
+  panelRef: RefObject<HTMLElement | null>
+  /** Distance kept between the anchor's bottom edge and the panel's top. */
+  gap: number
+  /** Distance kept between the panel and each viewport edge. */
+  margin: number
+}
+
+/**
+ * Track an anchor and return the panel's fixed coordinates.
+ * @param options - the open state, the two refs, and the gap/margin distances.
+ * @returns `left`/`top` for the panel, or `null` before the first measurement.
+ */
+export function useAnchoredPosition(options: AnchoredPositionOptions): CSSProperties | null {
+  const { open, anchorRef, panelRef, gap, margin } = options
+  const [position, setPosition] = useState<CSSProperties | null>(null)
+  useLayoutEffect(() => {
+    if (!open) {
+      setPosition(null)
+      return
+    }
+    const place = () => {
+      /* v8 ignore start -- geometry read from real layout: jsdom reports zero
+         offset sizes, so the positive-size clamp arms are exercised by browser
+         scenarios rather than unit tests. */
+      const rect = anchorRef.current?.getBoundingClientRect()
+      if (rect === undefined) return
+      const panel = panelRef.current
+      const width = panel?.offsetWidth ?? 0
+      const height = panel?.offsetHeight ?? 0
+      let left = rect.left
+      let top = rect.bottom + gap
+      if (width > 0) left = Math.min(Math.max(left, margin), window.innerWidth - width - margin)
+      if (height > 0) top = Math.min(Math.max(top, margin), window.innerHeight - height - margin)
+      /* v8 ignore stop */
+      setPosition({ left, top })
+    }
+    // The first run measures the panel in the same commit that opened it, so
+    // the clamp uses real dimensions before anything paints.
+    place()
+    window.addEventListener('scroll', place, true)
+    window.addEventListener('resize', place)
+    // The panel's own height changes without either event — a status line
+    // appearing inside it, or a `resize: vertical` textarea dragged taller —
+    // and a stale clamp would let a panel near the bottom edge cross the
+    // margin it is supposed to respect. The guard keeps the hook usable where
+    // `ResizeObserver` is absent, which is how jsdom runs.
+    const panel = panelRef.current
+    let observer: ResizeObserver | null = null
+    if (typeof ResizeObserver !== 'undefined' && panel !== null) {
+      observer = new ResizeObserver(place)
+      observer.observe(panel)
+    }
+    return () => {
+      observer?.disconnect()
+      window.removeEventListener('scroll', place, true)
+      window.removeEventListener('resize', place)
+    }
+  }, [open, anchorRef, panelRef, gap, margin])
+  return position
+}

packages/client/ui-primitives/tests/use-anchored-position.client.spec.tsx

@@ -0,0 +1,111 @@
+// @vitest-environment jsdom
+/**
+ * `useAnchoredPosition` wiring: a floating panel is placed from its anchor and
+ * keeps tracking it while open.
+ *
+ * The geometry itself needs real layout, which jsdom does not provide — the
+ * browser layout scenario in `apps/web/tests/message-feedback-layout.e2e.ts`
+ * owns that. What is asserted here is the wiring the clamp depends on: the
+ * listeners and the panel-size observer are attached while open and released on
+ * close, a size change replays the placement, and the hook still works where
+ * `ResizeObserver` does not exist.
+ */
+import { useRef } from 'react'
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import { cleanup, render } from '@testing-library/react'
+import { useAnchoredPosition } from '../src/useAnchoredPosition.ts'
+
+afterEach(() => {
+  cleanup()
+  vi.unstubAllGlobals()
+})
+
+/** One recorded `ResizeObserver` instance, so a test can drive its callback. */
+interface Recorded {
+  callback: ResizeObserverCallback
+  observed: Element[]
+  disconnected: boolean
+}
+
+/**
+ * Install a recording `ResizeObserver` double.
+ * @returns the list every constructed observer registers itself in.
+ */
+function stubResizeObserver(): Recorded[] {
+  const made: Recorded[] = []
+  vi.stubGlobal('ResizeObserver', class {
+    private readonly record: Recorded
+
+    constructor(callback: ResizeObserverCallback) {
+      this.record = { callback, observed: [], disconnected: false }
+      made.push(this.record)
+    }
+
+    observe(element: Element) { this.record.observed.push(element) }
+    disconnect() { this.record.disconnected = true }
+  })
+  return made
+}
+
+/**
+ * Host component that anchors a panel and reports the computed position.
+ * @param props - whether the panel is open.
+ * @returns the anchor and, while open, the panel carrying the position.
+ */
+function Host({ open }: { open: boolean }) {
+  const anchorRef = useRef<HTMLButtonElement>(null)
+  const panelRef = useRef<HTMLDivElement>(null)
+  const position = useAnchoredPosition({ open, anchorRef, panelRef, gap: 4, margin: 12 })
+  return (
+    <>
+      <button ref={anchorRef} type="button">anchor</button>
+      {open && <div ref={panelRef} data-testid="panel" style={position ?? { visibility: 'hidden' }} />}
+    </>
+  )
+}
+
+describe('useAnchoredPosition', () => {
+  it('observes the panel while open and disconnects when it closes', () => {
+    const made = stubResizeObserver()
+    const ui = render(<Host open />)
+
+    expect(made).toHaveLength(1)
+    expect(made[0]?.observed).toEqual([ui.getByTestId('panel')])
+    expect(made[0]?.disconnected).toBe(false)
+
+    ui.rerender(<Host open={false} />)
+
+    expect(made[0]?.disconnected).toBe(true)
+  })
+
+  it('replaces the panel when its own size changes', () => {
+    const made = stubResizeObserver()
+    render(<Host open />)
+    const before = made[0]?.callback
+    expect(before).toBeDefined()
+
+    // A status line appearing inside the panel, or a dragged textarea, changes
+    // the height without a scroll or resize event; the observer is the only
+    // thing that notices, so driving its callback must not throw.
+    expect(() => { before?.([], {} as ResizeObserver) }).not.toThrow()
+  })
+
+  it('still places the panel where ResizeObserver does not exist', () => {
+    // jsdom's own condition, and any host without the API: the hook must fall
+    // back to scroll and resize rather than fail at mount.
+    vi.stubGlobal('ResizeObserver', undefined)
+
+    expect(() => render(<Host open />)).not.toThrow()
+  })
+
+  it('attaches no listeners while the element is closed', () => {
+    const made = stubResizeObserver()
+    const add = vi.spyOn(window, 'addEventListener')
+
+    render(<Host open={false} />)
+
+    expect(made).toHaveLength(0)
+    expect(add.mock.calls.filter(([type]) => type === 'scroll' || type === 'resize')).toEqual([])
+    add.mockRestore()
+  })
+})

packages/client/ui-tool/README.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 packages/client/ui-tool/README.md
-README.md: 6e2bef2f5ad4b136510c3acbb8f8b8e83c4c9212
-README.zh.md: 169417747541db9e02cb552b175ca7c100420aeb
+README.md: b87236309c9bafe3e35d3d5977d56bd62a24de31
+README.zh.md: 3bae0b3f4cb3ad695371ec7a66fb531a935a4d2f

packages/client/ui-tool/README.md

@@ -28,7 +28,7 @@ ctx.slots.inject('tool.call.toolview', () =>
   }, BusinessToolRow))
 ```
 
-The owner payload is `ToolCallOwnerProps`: `callId`, `toolName`, the frozen `block`, optional `cwd`, and plain `openFile`/`inspect` callbacks. The registration receives the normal session slot runtime share. It does not receive React nodes, Runtime services, or root/subcall knowledge.
+The owner payload is `ToolCallOwnerProps`: `callId`, `toolName`, the frozen `block`, optional `cwd` and `home`, and plain `openFile`/`inspect` callbacks. Path summaries relativize to the session cwd first, then replace a leftover POSIX host home with `~`; `filePath` and Host open keep the authored filesystem path. The registration receives the normal session slot runtime share. It does not receive React nodes, Runtime services, or root/subcall knowledge.
 
 This package currently owns the generic fallback and the built-in shell/pwsh, read, write/edit, grep/glob, web, todo, question, and Code Dispatch presentations. `ui-skill` demonstrates a business-owned registration for `skill`.
 

packages/client/ui-tool/README.zh.md

@@ -28,7 +28,7 @@ ctx.slots.inject('tool.call.toolview', () =>
   }, BusinessToolRow))
 ```
 
-owner 载荷为 `ToolCallOwnerProps`：`callId`、`toolName`、冻结的 `block`、可选 `cwd`，以及普通的 `openFile`、`inspect` 回调。注册项会收到常规的会话 slot 运行时共享数据，但不会收到 React node、运行时服务或 root/subcall 知识。
+owner 载荷为 `ToolCallOwnerProps`：`callId`、`toolName`、冻结的 `block`、可选 `cwd` 与 `home`，以及普通的 `openFile`、`inspect` 回调。路径摘要先相对会话 cwd 缩短，再把剩余的 POSIX 宿主家目录写成 `~`；`filePath` 与 Host 打开仍使用作者给出的文件系统路径。注册项会收到常规的会话 slot 运行时共享数据，但不会收到 React node、运行时服务或 root/subcall 知识。
 
 本包当前拥有 generic fallback，以及 shell/pwsh、read、write/edit、grep/glob、web、todo、question 和 Code Dispatch 的内置展示。`ui-skill` 展示了业务包自行拥有的 `skill` 注册项。
 

packages/client/ui-tool/package.json

@@ -32,6 +32,7 @@
   "dsh": {
     "client": {
       "inject": [
+        "@deepseek-ai/dsh-client-connection",
         "@deepseek-ai/dsh-client-runtime",
         "@deepseek-ai/dsh-client-locale",
         "@deepseek-ai/dsh-client-ui-conversation"
@@ -50,6 +51,7 @@
   "peerDependencies": {
     "@deepseek-ai/cordis": "workspace:^",
     "@deepseek-ai/dsh-api-remotes": "workspace:^",
+    "@deepseek-ai/dsh-client-connection": "workspace:^",
     "@deepseek-ai/dsh-client-locale": "workspace:^",
     "@deepseek-ai/dsh-client-runtime": "workspace:^",
     "@deepseek-ai/dsh-client-ui-conversation": "workspace:^",

packages/client/ui-tool/src/client/apply.ts

@@ -1,4 +1,5 @@
 /** Register the Tool call tree, details renderer, and built-in atomic views. */
+import type { ConnectionHandle } from '@deepseek-ai/dsh-client-connection/client'
 import type { ClientContext } from '@deepseek-ai/dsh-client-runtime/client'
 import type {} from '@deepseek-ai/dsh-client-ui-conversation/client'
 import { ToolCallTree } from './tool/ToolCallTree.tsx'
@@ -12,14 +13,16 @@ import { searchToolview } from './tool/toolviews/search-row.tsx'
 import { todoToolview } from './tool/toolviews/todo-row.tsx'
 import { webToolview } from './tool/toolviews/web-row.tsx'
 
-/** Required service: the slot registry that owns both Tool render seats. */
-export const inject = ['slots']
+/** Required services: the slot registry and the Host description used for POSIX `~`. */
+export const inject = ['slots', 'connection']
 
 /**
  * Mount the whole-Tool renderers and built-in atomic Tool registrations.
  * @param ctx - Client root context.
  */
 export function apply(ctx: ClientContext): void {
+  const connection = ctx.get('connection') as ConnectionHandle
+  const toolInject = () => ({ hooks: { hostDescription: connection.hostDescription } })
   ctx.slots.inject('conversation.chat.node', () => ctx.slots.register({
     name: 'conversation.chat.node',
     key: 'tool-call',
@@ -27,11 +30,13 @@ export function apply(ctx: ClientContext): void {
     children: {
       'tool.call.toolview': { kind: 'keyed', scope: 'session' },
     },
+    inject: toolInject,
   }, ToolCallTree))
 
   ctx.slots.inject('conversation.details.tool', () => ctx.slots.register({
     name: 'conversation.details.tool',
     locale: NS,
+    inject: toolInject,
   }, ToolDetails))
 
   ctx.plugin(bashToolviewSample)

packages/client/ui-tool/src/client/contract/slots.ts

@@ -1,5 +1,6 @@
 /** Tool UI slot declarations and their composed component props. */
-import type { PropsLocale, PropsRenderSlots, PropsRuntime } from '@deepseek-ai/dsh-client-ui-slots'
+import type { HostDescriptionSource } from '@deepseek-ai/dsh-client-connection/client'
+import type { InjectFace, PropsLocale, PropsRenderSlots, PropsRuntime } from '@deepseek-ai/dsh-client-ui-slots'
 import type { ToolCallBlock } from '@deepseek-ai/dsh-client-runtime/client'
 import type {} from '@deepseek-ai/dsh-client-ui-conversation/client'
 import type {} from '@deepseek-ai/dsh-client-locale/client'
@@ -34,6 +35,8 @@ export interface ToolCallOwnerProps {
   block: ToolCallBlock
   /** Session workspace root for relative summaries. */
   cwd?: string | undefined
+  /** Host account home; POSIX home-rooted summaries display as `~`. */
+  home?: string | undefined
   /** Open a Tool argument path through the Host. */
   openFile: (path: string) => void
   /** Inspect this call in the trajectory view when available. */
@@ -43,10 +46,21 @@ export interface ToolCallOwnerProps {
 /** Full props of a registered atomic Tool view. */
 export type ToolCallViewProps = PropsRuntime<'tool.call.toolview'>
 
+/** Injected Host description for POSIX home-path display. */
+export type ToolHostDescriptionInjected = {
+  hooks: {
+    /** Current generation's Host description, bound by the slot renderer. */
+    hostDescription: HostDescriptionSource
+  }
+}
+
 /** Full props of the Tool call-tree renderer registered as a `tool-call` Chat Node. */
 export type ToolTreeProps = PropsRuntime<'conversation.chat.node', 'tool-call'>
   & PropsRenderSlots<'tool.call.toolview'>
   & PropsLocale<'conversation'>
+  & InjectFace<ToolHostDescriptionInjected>
 
 /** Full props of the selected Tool output renderer in the details panel. */
-export type ToolDetailsProps = PropsRuntime<'conversation.details.tool'> & PropsLocale<'conversation'>
+export type ToolDetailsProps = PropsRuntime<'conversation.details.tool'>
+  & PropsLocale<'conversation'>
+  & InjectFace<ToolHostDescriptionInjected>

packages/client/ui-tool/src/client/index.ts

@@ -1,3 +1,5 @@
 /** Browser Tool plugin: whole-call composition and keyed atomic Tool views. */
 export { apply, inject } from './apply.ts'
-export type { ToolCallOwnerProps, ToolCallViewProps, ToolDetailsProps, ToolTreeProps } from './contract/slots.ts'
+export type {
+  ToolCallOwnerProps, ToolCallViewProps, ToolDetailsProps, ToolHostDescriptionInjected, ToolTreeProps,
+} from './contract/slots.ts'

packages/client/ui-tool/src/client/tool/ToolCallTree.tsx

@@ -12,12 +12,13 @@ function callName(node: ToolCallBlock): string {
 
 /** One atomic call dispatched through the Tool-owned keyed slot. */
 const ToolCall = memo(function ToolCall({
-  renderSlot, callId, toolName, block, openFile, selected, cwd, inspectCall, t, children,
+  renderSlot, callId, toolName, block, openFile, selected, cwd, home, inspectCall, t, children,
 }: Pick<ToolTreeProps, 'renderSlot' | 'openFile' | 'cwd' | 'inspectCall' | 't'> & {
   callId: string
   toolName: string
   block: ToolCallBlock
   selected: boolean
+  home?: string | undefined
   children?: ReactNode
 }) {
   const owner: ToolCallOwnerProps = useMemo(() => ({
@@ -26,8 +27,9 @@ const ToolCall = memo(function ToolCall({
     block,
     openFile,
     cwd,
+    home,
     inspect: () => { inspectCall(callId) },
-  }), [callId, toolName, block, openFile, cwd, inspectCall])
+  }), [callId, toolName, block, openFile, cwd, home, inspectCall])
   return (
     <div
       className={css.callRow}
@@ -45,9 +47,10 @@ const ToolCall = memo(function ToolCall({
 })
 
 const ToolCallBranch = memo(function ToolCallBranch({
-  renderSlot, block, selectedCallId, cwd, openFile, inspectCall, t,
+  renderSlot, block, selectedCallId, cwd, home, openFile, inspectCall, t,
 }: Pick<ToolTreeProps, 'renderSlot' | 'selectedCallId' | 'cwd' | 'openFile' | 'inspectCall' | 't'> & {
   block: ToolCallBlock
+  home?: string | undefined
 }) {
   return (
     <ToolCall
@@ -58,6 +61,7 @@ const ToolCallBranch = memo(function ToolCallBranch({
       openFile={openFile}
       selected={block.callId === selectedCallId}
       cwd={cwd}
+      home={home}
       inspectCall={inspectCall}
       t={t}
     >
@@ -70,6 +74,7 @@ const ToolCallBranch = memo(function ToolCallBranch({
               block={child}
               selectedCallId={selectedCallId}
               cwd={cwd}
+              home={home}
               openFile={openFile}
               inspectCall={inspectCall}
               t={t}
@@ -88,8 +93,9 @@ const ToolCallBranch = memo(function ToolCallBranch({
  * @returns the Tool call tree.
  */
 export function ToolCallTree({
-  renderSlot, node, selectedCallId, cwd, openFile, inspectCall, t,
+  renderSlot, node, selectedCallId, cwd, openFile, inspectCall, useHostDescription, t,
 }: ToolTreeProps) {
+  const home = useHostDescription(description => description?.home)
   const block = node.data.root
   return (
     <ToolCallBranch
@@ -97,6 +103,7 @@ export function ToolCallTree({
       block={block}
       selectedCallId={selectedCallId}
       cwd={cwd}
+      home={home}
       openFile={openFile}
       inspectCall={inspectCall}
       t={t}

packages/client/ui-tool/src/client/tool/ToolDetails.tsx

@@ -9,20 +9,16 @@ import { resultText } from './models/tool-call-model.ts'
 import { webCardModel } from './models/web-card-model.ts'
 import css from './ToolDetails.module.css'
 
-/** Pure details-body inputs; framework session seats stay at the slot boundary. */
-interface ToolDetailsContentProps {
-  block: ToolDetailsProps['block']
-  cwd?: ToolDetailsProps['cwd']
-  t: ToolDetailsProps['t']
-}
-
 /**
  * Render the selected Tool call's structured output when its presentation
  * intent is known, otherwise preserve the flattened result text.
- * @param props - selected call slice, workspace root, and locale seat.
+ * @param props - selected call slice, workspace root, host home, and locale seat.
  * @returns the details output body.
  */
-export function ToolDetails({ block, cwd, t }: ToolDetailsContentProps) {
+export function ToolDetails({
+  block, cwd, useHostDescription, t,
+}: Pick<ToolDetailsProps, 'block' | 'cwd' | 'useHostDescription' | 't'>) {
+  const home = useHostDescription(description => description?.home)
   const terminal = terminalCardModel(block, cwd)
   if (terminal !== null) {
     return (
@@ -34,7 +30,7 @@ export function ToolDetails({ block, cwd, t }: ToolDetailsContentProps) {
       </>
     )
   }
-  const read = readCardModel(block, cwd)
+  const read = readCardModel(block, cwd, home)
   if (read !== null) return <ReadBlock {...read} className={css.read} />
   const diff = diffCardModel(block)
   if (diff !== null) return <DiffBlock {...diff.card} className={css.cardBody} />

packages/client/ui-tool/src/client/tool/models/read-card-model.ts

@@ -13,6 +13,7 @@
  * until the result arrives.
  * @module
  */
+import { abbreviateHomePath } from '@deepseek-ai/dsh-client-runtime/client'
 import type { ReadBlockLine, ReadBlockProps } from '@deepseek-ai/dsh-client-ui-primitives'
 import { relativizeToCwd, type ToolCallBlock } from './tool-call-model.ts'
 
@@ -52,14 +53,15 @@ export type ReadCardModel = Pick<ReadBlockProps, 'label' | 'lines' | 'totalLines
  *
  * The label is the read view's `title` when the tool supplied one (the
  * presentation contract's replacement-title rule), otherwise the file path
- * relativized to the session workspace so a workspace-rooted absolute path
- * displays the same short form the row summary shows.
+ * shortened the same way the row summary is: workspace-relative first, then
+ * POSIX `~` for a leftover host-home path.
  * @param block - RunningToolCall or ToolResultNode off the snapshot caches.
  * @param sessionCwd - the session workspace root; a workspace-rooted absolute
  *   path label displays relative to it. Absent leaves the path as authored.
+ * @param home - host account home; a leftover POSIX home path displays as `~`.
  * @returns the read-card props, or null for the generic path.
  */
-export function readCardModel(block: ToolCallBlock, sessionCwd?: string): ReadCardModel | null {
+export function readCardModel(block: ToolCallBlock, sessionCwd?: string, home?: string): ReadCardModel | null {
   // Running has no result view; a read carries no content until execute returns.
   if (!('kind' in block)) return null
   const result = block.resultView?.card === 'read' ? block.resultView : null
@@ -68,7 +70,7 @@ export function readCardModel(block: ToolCallBlock, sessionCwd?: string): ReadCa
   // shape so the card never holds a reference into the runtime's cache.
   const lines: ReadBlockLine[] = result.lines.map(line => ({ number: line.number, text: line.text }))
   return {
-    label: result.title ?? relativizeToCwd(result.path, sessionCwd),
+    label: result.title ?? abbreviateHomePath(relativizeToCwd(result.path, sessionCwd), home),
     lines,
     totalLines: result.totalLines,
     lang: result.lang,

packages/client/ui-tool/src/client/tool/models/tool-call-model.ts

@@ -9,6 +9,7 @@
 // The block union's defining home is runtime (fold-product types); this
 // contract only forwards it (type-definition authority stays with the layer
 // that produces the values).
+import { abbreviateHomePath } from '@deepseek-ai/dsh-client-runtime/client'
 import type { ToolCallBlock, ToolResultNode } from '@deepseek-ai/dsh-client-runtime/client'
 
 export type { ToolCallBlock } from '@deepseek-ai/dsh-client-runtime/client'
@@ -210,16 +211,19 @@ function deriveBody(variant: ToolRowVariant, argsRaw: string): string | null {
  * @param toolName - wire tool name (dispatch-supplied; survives windowless results).
  * @param block - RunningToolCall or ToolResultNode off the snapshot caches.
  * @param cwd - session workspace root; workspace-rooted path summaries display relative to it.
+ * @param home - host account home; a leftover POSIX home path displays as `~`.
  * @returns the row model.
  */
-export function toolRowModel(toolName: string, block: ToolCallBlock, cwd?: string): ToolRowModel {
+export function toolRowModel(toolName: string, block: ToolCallBlock, cwd?: string, home?: string): ToolRowModel {
   const variant = classifyTool(toolName)
   const done = 'kind' in block
   const argsRaw = (done ? block.call?.argsRaw : block.argsRaw) ?? ''
   const state: ToolRowState = !done ? 'running'
     : block.error?.code === 'interrupted' ? 'stopped'
       : block.isError ? 'error' : 'ok'
-  const base = argsRaw === '' ? block.callId : relativizeToCwd(deriveSummary(variant, argsRaw), cwd)
+  const base = argsRaw === ''
+    ? block.callId
+    : abbreviateHomePath(relativizeToCwd(deriveSummary(variant, argsRaw), cwd), home)
   const toolTitle = TOOL_TITLES[toolName]
   // Others keeps the static "Tool call" title (figma literal); the real tool
   // name rides the mutable summary slot unless the tool owns a specific title.

packages/client/ui-tool/src/client/tool/toolviews/GenericToolCard.tsx

@@ -33,10 +33,10 @@ export interface GenericToolCardProps extends ToolCallOwnerProps {
   t: ToolTreeProps['t']
 }
 
-export function GenericToolCard({ toolName, block, cwd, openFile, inspect, t }: GenericToolCardProps) {
-  const model = toolRowModel(toolName, block, cwd)
+export function GenericToolCard({ toolName, block, cwd, home, openFile, inspect, t }: GenericToolCardProps) {
+  const model = toolRowModel(toolName, block, cwd, home)
   const terminal = terminalCardModel(block, cwd)
-  const read = readCardModel(block, cwd)
+  const read = readCardModel(block, cwd, home)
   const diff = diffCardModel(block)
   const search = searchCardModel(block)
   const web = webCardModel(block)

packages/client/ui-tool/src/client/tool/toolviews/file-mutation-row.tsx

@@ -29,8 +29,8 @@ type FileMutationRowProps = ToolCallViewProps & PropsLocale<'conversation'>
  * model-facing error text through its Output section and its first line in the
  * collapsed summary instead.
  */
-export function FileMutationRow({ toolName, block, cwd, openFile, inspect, t }: FileMutationRowProps) {
-  const model = toolRowModel(toolName, block, cwd)
+export function FileMutationRow({ toolName, block, cwd, home, openFile, inspect, t }: FileMutationRowProps) {
+  const model = toolRowModel(toolName, block, cwd, home)
   const diff = diffCardModel(block)
   return (
     <ToolRow

packages/client/ui-tool/src/client/tool/toolviews/read-row.tsx

@@ -24,9 +24,9 @@ type ReadRowProps = ToolCallViewProps & PropsLocale<'conversation'>
  * read card as the row's collapsed-by-default card body. The summary path is an
  * openable host link when the row names a single file.
  */
-export function ReadRow({ toolName, block, cwd, openFile, inspect, t }: ReadRowProps) {
-  const model = toolRowModel(toolName, block, cwd)
-  const read = readCardModel(block, cwd)
+export function ReadRow({ toolName, block, cwd, home, openFile, inspect, t }: ReadRowProps) {
+  const model = toolRowModel(toolName, block, cwd, home)
+  const read = readCardModel(block, cwd, home)
   return (
     <ToolRow
       t={t}