Merge origin/master into codex/dsh-badge-plugin

.agents/notes/implemented/architecture/2026-07-19-package-invariant-runtime-contracts.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/2026-07-19-package-invariant-runtime-contracts.md
-2026-07-19-package-invariant-runtime-contracts.md: 40d152b2320ac65f9ea7d8732b1a667236d2780a
-2026-07-19-package-invariant-runtime-contracts.zh.md: a734b1a3deb739c214d1c4c2565fc697b5e4b89b
+2026-07-19-package-invariant-runtime-contracts.md: 86d86f69b606c348e85d1ae654b6b35c4326985a
+2026-07-19-package-invariant-runtime-contracts.zh.md: 386a294b577561d08493a3b2925910ee5ba971b4

.agents/notes/implemented/architecture/2026-07-19-package-invariant-runtime-contracts.md

@@ -50,7 +50,7 @@ The current 103-package workspace has 21 executable companions and 82 justified
 | `dsh-user-approval` | Approval asked/decided records pair by call and use valid outcomes and policies. |
 | `dsh-workflow` | Workflow and child-agent start/end events preserve run metadata, identity, outcome, count, and error relations. |
 | `dsh-tasks` | Current and terminal task snapshots preserve id/kind, owner, status, and timestamp relationships. |
-| `dsh-tool-todo` | Durable whole-list snapshots use unique trimmed items, closed statuses, and at most one active item. |
+| `dsh-tool-todo` | Durable whole-list snapshots use unique trimmed items and closed statuses. |
 | `dsh-time-context` | Plugin-attributed clock readings agree with the session's open turn, next pre-step position, and elapsed baseline; rendered time parses and does not postdate its event. |
 
 Session-backed companions validate existing durable events when they load, using the prefix preceding each candidate where the relationship depends on event order. Other checks observe the authoritative live event boundary or mutable service result. Validation runs before publication where accepting an invalid event would otherwise commit bad state.

.agents/notes/implemented/architecture/2026-07-19-package-invariant-runtime-contracts.zh.md

@@ -50,7 +50,7 @@ Status: implemented
 | `dsh-user-approval` | approval asked/decided 记录按 call 配对，并使用有效 outcome 和 policy。 |
 | `dsh-workflow` | 工作流和 child-agent start/end 事件保持 run metadata、身份、outcome、数量和 error 关系。 |
 | `dsh-tasks` | 当前与终态 task 快照保持 id/kind、owner、status 和 timestamp 关系。 |
-| `dsh-tool-todo` | 持久化全量快照使用唯一且已 trim 的条目、封闭 status，并且最多有一个活动条目。 |
+| `dsh-tool-todo` | 持久化全量快照使用唯一且已 trim 的条目和封闭 status。 |
 | `dsh-time-context` | 标注插件来源的时钟 reading 必须匹配会话当前打开的轮次、下一个步骤开始前的位置和 elapsed baseline；渲染时间必须可解析，且不得晚于对应事件。 |
 
 基于会话的 companion 在加载时验证已有持久化事件；关系依赖事件顺序时，会使用每个候选事件之前的事件前缀。其他检查观测权威实时事件边界或可变服务结果。如果接受无效事件会提交错误状态，验证就在发布前执行。

.agents/notes/implemented/architecture/2026-07-27-dispose-ladder-to-consumer.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/2026-07-27-dispose-ladder-to-consumer.md
-2026-07-27-dispose-ladder-to-consumer.md: 97b551ff509e3b424f6bf5725939cf54acc961a7
-2026-07-27-dispose-ladder-to-consumer.zh.md: 7fff744e64109549a65d4f5bb17ff2d6ddfc6888
+2026-07-27-dispose-ladder-to-consumer.md: e9af88e8e7ef962213a74e96a249241cbe8d5994
+2026-07-27-dispose-ladder-to-consumer.zh.md: 89f8e107c56d42787c59bc6f8fa8fc7b3ef73208

.agents/notes/implemented/architecture/2026-07-27-dispose-ladder-to-consumer.md

@@ -10,7 +10,7 @@ English | [中文](2026-07-27-dispose-ladder-to-consumer.zh.md)
 
 ## Decision
 
-The ladder moves to its one consumer. `dsh-subagent-acp` owns `disposeAcpChild(child, eofGraceMs, graceMs)`, built entirely on the seam's public verbs: close `stdin`, bound a `waitForExit` on `eofGraceMs`, then `terminate()` (whose SIGTERM→spec-grace→SIGKILL escalation already encodes the signal tiers), then a final bounded whole-tree wait that throws if survivors remain. The seam keeps `kill`/`terminate`/`waitForExit` — mechanisms, not policy — and `waitForExit(signal?)` is exactly the quiescence probe a consumer ladder needs to hold each tier on real tree exit. `dsh-subprocess-local` drops its `dsh-timeout` dependency; the seam's handle loses one method and one exported interface.
+The ladder moves to its one consumer. `dsh-subagent-acp` owns `disposeAcpChild(child, eofGraceMs)`, built entirely on the seam's public verbs: close `stdin`, bound a `waitForExit` on `eofGraceMs`, then call `terminate()`, whose SIGTERM→spec-grace→SIGKILL escalation already owns the signal timer, and await an unbounded `waitForExit()` for the subprocess owner's whole-tree exit proof. The seam keeps `kill`/`terminate`/`waitForExit` — mechanisms, not policy — and `waitForExit(signal?)` is exactly the quiescence probe a consumer ladder needs to hold the cooperative tier on real tree exit without deriving another timer from the termination grace. The seam's handle loses one method and one exported interface.
 
 ## Alternatives considered
 
@@ -20,4 +20,4 @@ The ladder moves to its one consumer. `dsh-subagent-acp` owns `disposeAcpChild(c
 
 ## Consequences
 
-Bought: the seam is one method and one type smaller; implementations owe four verbs and no teardown policy; `dsh-subprocess-local` loses a dependency; the ladder's tier windows live beside the config fields that tune them. Cost: a future backend wanting EOF-first teardown writes ~20 lines against the verbs (or lifts the ACP helper); the ladder's tier-tier tests moved from the seam suite to the ACP suite, and the seam suite pins the verbs the ladder composes (bounded `waitForExit` false-then-true across an escalation) instead of the composed policy.
+Bought: the seam is one method and one type smaller; implementations owe four verbs and no teardown policy; the cooperative EOF window lives beside the ACP config field that tunes it, while the subprocess owner alone owns the termination window and final join. Cost: a future backend wanting EOF-first teardown writes ~20 lines against the verbs (or lifts the ACP helper); the ladder's tier tests live in the ACP suite, and the seam suite pins the verbs the ladder composes (bounded `waitForExit` false before escalation and an unbounded whole-tree join after it) instead of the composed policy.

.agents/notes/implemented/architecture/2026-07-27-dispose-ladder-to-consumer.zh.md

@@ -10,7 +10,7 @@ Status: implemented
 
 ## 决策
 
-阶梯移入其唯一消费方。`dsh-subagent-acp` 拥有 `disposeAcpChild(child, eofGraceMs, graceMs)`，完全构建在 seam 的公开动词之上：关闭 `stdin`，以 `eofGraceMs` 约束一次 `waitForExit`，随后 `terminate()`（其 SIGTERM→spec 宽限期→SIGKILL 升级已编码了信号层级），最后进行有界的整树等待，若仍有存活进程则抛出。seam 保留 `kill`／`terminate`／`waitForExit`——机制而非策略——而 `waitForExit(signal?)` 恰是消费方阶梯在每一层确认进程树真正退出所需的完全停稳探针。`dsh-subprocess-local` 卸下 `dsh-timeout` 依赖；seam 的句柄少了一个方法和一个导出接口。
+阶梯移入其唯一消费方。`dsh-subagent-acp` 拥有 `disposeAcpChild(child, eofGraceMs)`，完全构建在 seam 的公开动词之上：关闭 `stdin`，以 `eofGraceMs` 约束一次 `waitForExit`，随后调用 `terminate()`（其 SIGTERM→spec 宽限期→SIGKILL 升级已拥有信号定时器），再无界等待 `waitForExit()`，由子进程责任方证明整棵进程树已经退出。seam 保留 `kill`／`terminate`／`waitForExit`——机制而非策略——而 `waitForExit(signal?)` 恰是消费方阶梯在协作层确认进程树真正退出所需的停稳探针，无需从终止宽限期再派生一个定时器。seam 的句柄少了一个方法和一个导出接口。
 
 ## 曾考虑的替代方案
 
@@ -20,4 +20,4 @@ Status: implemented
 
 ## 后果
 
-换来的是：seam 少了一个方法和一个类型；实现只需提供四个动词，无需提供拆卸策略；`dsh-subprocess-local` 少了一个依赖；阶梯的层级时间窗与调节它们的配置字段住在一起。代价：未来想要 EOF 打头拆卸的后端需针对这些动词写约 20 行（或直接搬 ACP 的辅助函数）；阶梯的层级测试从 seam 套件移入 ACP 套件，seam 套件转而钉住阶梯所组合的动词（升级前后有界 `waitForExit` 先假后真），而非组合后的策略。
+买到的：seam 少了一个方法和一个类型；实现只欠四个动词，不欠拆卸策略；协作式 EOF 时间窗与调节它的 ACP 配置字段住在一起，而终止时间窗与最终的整树退出等待仅由子进程责任方拥有。代价：未来想要 EOF 打头拆卸的后端需针对这些动词写约 20 行（或直接搬 ACP 的辅助函数）；阶梯的层级测试位于 ACP 套件，seam 套件转而钉住阶梯所组合的动词（升级前有界 `waitForExit` 返回假，升级后无界等待整棵进程树退出），而非组合后的策略。

.agents/notes/implemented/architecture/2026-07-30-web-config-plane.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/2026-07-30-web-config-plane.md
-2026-07-30-web-config-plane.md: 5225460be1d66b85a05ff2fd5ae2826b0e6c41d7
-2026-07-30-web-config-plane.zh.md: 53a21ddf31640d963c413e1793276de694547311
+2026-07-30-web-config-plane.md: 0b18cee414df23a2ed8a8b43b76dc06403804691
+2026-07-30-web-config-plane.zh.md: e70c2a47970f943e49393b099c4fcea58dc0fbdc

.agents/notes/implemented/architecture/2026-07-30-web-config-plane.md

@@ -16,13 +16,13 @@ PR1 made LLM adapter configuration restart-free at the seam, but the only writer
 
 **`describe()` grows layers and structural secret redaction.** `SettingsDescriptor` carries `base`/`user` beside the effective value, so the form marks "overridden" by presence in the user layer, not value inequality (an override *equal* to the base is still an override). `describe({ redactSecrets: true })` — mandatory at every wire face — strips `role('secret')` subtrees from all three layers via a pure structural walk of the schema (object/dict/array containers; a secret-role subtree is one opaque leaf) and enumerates the stripped slots as `{path, set}`, so a page can render write-only inputs without ever receiving a value.
 
-**The Host identifies and opens the local settings document.** The settings seam exposes optional `documentPath` provider metadata and a `prepareDocument()` operation; `settings-local` returns its fully resolved custom or `$DSH_HOME/settings.yaml` filename and exclusively creates an absent empty document with owner-only permissions, while non-file providers retain the base `undefined`. The loopback-only `settings.describe` response carries only the boolean `hasDocument` capability beside the redacted namespace views. `ui-settings-general` registers a `settings.action` entry only on loopback pages, shows it only after the metadata confirms that a provider-owned local document can be prepared, and invokes pathless `settings.openDocument`; the Host resolves the provider path again before a text-document handoff (`open -t` on macOS so an arbitrary YAML file association cannot redirect the gesture, `xdg-open` on Linux, and `Invoke-Item` on Windows). Generic workspace paths retain the existing default-application handoff. The browser neither derives `$DSH_HOME` nor receives a filesystem target; remote pages make no privileged settings read for this action.
+**The Host identifies and opens the local settings document.** The settings seam exposes optional `documentPath` provider metadata and a `prepareDocument()` operation; `settings-local` returns its fully resolved custom or `$DSH_HOME/settings.yaml` filename and exclusively creates an absent empty document with owner-only permissions, while non-file providers retain the base `undefined`. The loopback-only `settings.describe` response carries only the boolean `hasDocument` capability beside the redacted namespace views. `ui-settings-general` registers a `settings.action` entry only on loopback pages, shows it only after the metadata confirms that a provider-owned local document can be prepared, and invokes pathless `settings.openDocument`; the Host resolves the provider path again before a text-document handoff (`open -t` on macOS so an arbitrary YAML file association cannot redirect the gesture, `xdg-open` on desktop Linux, `Invoke-Item` on Windows, and `wslpath -w` followed by that Windows handoff on WSL). Generic workspace paths retain the existing default-application handoff. The browser neither derives `$DSH_HOME` nor receives a filesystem target; remote pages make no privileged settings read for this action.
 
 **The llm seam declares configurability and announces topology.** `registerConfigurableProviders()` is an all-or-nothing, fiber-scoped directory of `{provider, displayName, settingsNs, settingsPath}` — the addressing a config page needs to open the right settings subtree for a route that may not exist yet; `listConfigurableProviders()` merges with live routes in the wire handler so undeclared live routes still report active. The zero-payload `'llm/adapters-updated'` event fires from all four registration/unregistration commit points with contained listener dispatch (INVARIANT rethrow), following the settings/commands precedent. `llm-deepseek`'s route renamed to `deepseek-official` because the pi-ai catalog legitimately owns `deepseek` as an aggregator entry; pre-release stance, no alias.
 
 **A hand-written editor over a schema model layer.** `dsh-client-schema-form` rehydrates the wire's `toJSON()` envelope into live schemastery nodes for validation, path resolution, and immutable draft editing — but no generic rendering: the first cut shipped a full schema-driven form renderer, and the resulting page was an unstyled schema dump (every advanced field flattened onto the card, raw field names as labels, the `retryPolicy` unsupported-fallback in the main flow). The user chose the hand-written direction over adding a hint/grouping system, and a second round removed the reference input entirely: the card's primary field is one **API key** input, a whole-section provider without a configured key opens as its setup card, and the collapsed 自定义设置 fold carries the curated per-family extras (`baseURL` for both families, `reasoningEffort` for deepseek / `reasoning` for pi-ai, plus direct DeepSeek model rows with `id`, `name`, and `contextWindow`). Existing model fields outside that visible set survive array edits; retry policy, timeouts, and other fields remain owned by `settings.yaml`. Validation still runs the rehydrated schema before writing, while adapter-specific checks reject catalog invariants that the serialized schema cannot express. The card's colors resolve through the `--dsw-alias-*` design tokens; it had named `--border`/`--surface`/`--text-*`, which nothing in this app defines, so it rendered their light-mode fallbacks and stayed light under the dark theme. The model catalog takes the row shape the pi-ai provider form introduces: one bordered entry per model, id and display name on the row, and the capacities behind the row's own disclosure, so the two editors read as one design rather than diverging once both land. Every field keeps the indexed `aria-label` that names it. Both capacities are text fields reading a decimal `K`/`M` suffix (`1M` is 1000K, matching how capacities are quoted) and storing the plain count: a field holds the typed text while it has focus, because re-deriving it from the parsed count on every keystroke would rewrite `1000` to `1K` mid-word, and text that does not parse stays on screen so the save-time rejection names a row the user can still see. The shared class names carry this file's token spellings, not that branch's: `--dsw-alias-border-subtle`, `--dsw-alias-text-tertiary`, and `--dsw-alias-text-primary` are undeclared, so they resolve to the light-mode literals in their fallback slots — the defect this section was moved off. A styles test now rejects any `--dsw-*` name the token sheet does not declare, so the next editor to name one fails rather than shipping a light-only surface.
 
-**The Models page is a three-domain join with seam-shaped apply semantics.** Rows are configured providers; the add card's select is the dormant directory remainder. Route liveness still gates readiness and invalidates the join, but the page does not render it as provider status because configuration presence and runtime availability are distinct. The key path stays reference-shaped without ever showing a reference: a typed key stores **write-only** through `credentials.set` under the profile's `apiKeyEnv`, deriving `<ROUTE>_API_KEY` when none exists (the pi-ai profile records the derivation), so `settings.yaml` never carries a key value. Profile edits and removals land as minimal path-addressed `settings.mutate` operations against the redacted user section, which never names a secret the page did not receive. Removing a user-layer provider first opens a localized model-provider confirmation dialog; cancellation, its close button, and its mask leave the profile untouched, while the destructive confirmation submits the single unset and blocks duplicate submission until it settles. DeepSeek's model list is array-replace configuration: inherited effective rows remain visible until the first edit materializes the complete list in the user layer, and reset unsets the list override.
+**The Models page is a three-domain join with seam-shaped apply semantics.** Rows are configured providers; the add card's select is the dormant directory remainder. Route liveness still gates readiness and invalidates the join, but the page does not render it as provider status because configuration presence and runtime availability are distinct. The key path stays reference-shaped without ever showing a reference: a typed key stores **write-only** through `credentials.set` under the profile's `apiKeyEnv`, deriving `<ROUTE>_API_KEY` when none exists (the pi-ai profile records the derivation only when a key is entered), so `settings.yaml` never carries a key value; a blank pi-ai key materializes a reference-free profile and preserves provider-native authentication. Profile edits and removals land as minimal path-addressed `settings.mutate` operations against the redacted user section, which never names a secret the page did not receive. Removing a user-layer provider first opens a localized confirmation dialog whose row actions, title, description, and final action identify the same provider; confirmation removes an exact configured+writable derived credential before the profile, while custom, environment, and unidentified targets remain untouched. Both stages are idempotent and a partial failure stays in the dialog for retry. DeepSeek's model list is array-replace configuration: inherited effective rows remain visible until the first edit materializes the complete list in the user layer, and reset unsets the list override. The partial-commit and credential-ownership rationale lives in the [provider credential lifecycle note](../bug-fix/2026-08-06-provider-credential-lifecycle.md).
 
 ## Alternatives considered
 
@@ -36,4 +36,4 @@ PR1 made LLM adapter configuration restart-free at the seam, but the only writer
 
 ## Consequences
 
-The whole loop is pinned keyless in the browser lane (`apps/web/tests/models-settings.e2e.ts`): the add card offers the dormant pi-ai catalog, adding `minimax-cn` with a typed key writes the reference-only profile into `settings.yaml`, stores the value into the harness home's `.env` under the derived `MINIMAX_CN_API_KEY`, registers the route live on the topology frame, and the customized fold merges `reasoning` beside the reference — zero model calls, ARIA goldens for the add-card, configured, and delete-confirmation states, plus a scaffold `harnessHome` so tests never touch a real `~/.dsh` (the provider under test is one whose derived reference cannot collide with a developer's exported keys). The settings-shell scenario intercepts the pathless native intent; seam, provider, wire, React, and native-opener tests separately pin provider absence, custom-path resolution, absent-file materialization, owner-only permissions, hidden remote/unavailable states, duplicate-click collapse, localized failure, macOS text-editor dispatch, and Linux/Windows desktop dispatch. The removal scenario proves cancellation leaves the profile intact, confirmation removes it, and the intentionally retained credential survives. The DeepSeek onboarding fixture edits the default catalog into a user-owned list, persists an arbitrary model id/name/context window, removes the active row, and observes the model selector's empty-selection fallback. The rename touched 239 files (fixtures, goldens, docs, python) in one commit with no compatibility alias. The renderer replacement cost one commit and no wire change: apply semantics, redaction, and the directory join were renderer-agnostic all along. Deferred: a per-row models preview (the picker already lists models), a page address for live routes that never declared configurability, and explicit removal of a provider's retained credential.
+The whole loop is pinned keyless in the browser lane (`apps/web/tests/models-settings.e2e.ts`): the add card offers the dormant pi-ai catalog, adding `minimax-cn` with a typed key writes the reference-only profile into `settings.yaml`, stores the value into the harness home's `.env` under the derived `MINIMAX_CN_API_KEY`, registers the route live on the topology frame, and the customized fold merges `reasoning` beside the reference — zero model calls, ARIA goldens for the add-card, configured, and identified delete-confirmation states, plus a scaffold `harnessHome` so tests never touch a real `~/.dsh` (the provider under test is one whose derived reference cannot collide with a developer's exported keys). The settings-shell scenario intercepts the pathless native intent; seam, provider, wire, React, and native-opener tests separately pin provider absence, custom-path resolution, absent-file materialization, owner-only permissions, hidden remote/unavailable states, duplicate-click collapse, localized failure, macOS text-editor dispatch, and Linux/Windows desktop dispatch. The removal scenario proves cancellation leaves both profile and key intact, then confirmation removes both the profile and its identified managed credential. The DeepSeek onboarding fixture edits the default catalog into a user-owned list, persists an arbitrary model id/name/context window, removes the active row, and observes the model selector's empty-selection fallback. The rename touched 239 files (fixtures, goldens, docs, python) in one commit with no compatibility alias. The renderer replacement cost one commit and no wire change: apply semantics, redaction, and the directory join were renderer-agnostic all along. Deferred: a per-row models preview (the picker already lists models) and a page address for live routes that never declared configurability.

.agents/notes/implemented/architecture/2026-07-30-web-config-plane.zh.md

@@ -16,13 +16,13 @@ PR1 让 LLM（大语言模型）适配器配置在 seam 层面免重启，但唯
 
 **`describe()` 增加分层与结构化 secret 脱敏。**`SettingsDescriptor` 在生效值之外携带 `base`/`user`，表单据此按「字段是否出现在用户层」来标记「已覆盖」，而非按值是否不等（与 base *相等*的覆盖仍然是覆盖）。`describe({ redactSecrets: true })`——在每个 wire 面都强制启用——经由对 schema 的纯结构遍历（object/dict/array 容器；secret 角色子树整体是一个不透明叶节点）从全部三层剥除 `role('secret')` 子树，并把剥除的槽位枚举为 `{path, set}`，页面因此不必收到任何值就能渲染只写输入框。
 
-**Host 识别并打开本地设置文档。** settings seam 暴露可选的 `documentPath` 提供方元数据和 `prepareDocument()` 操作；`settings-local` 返回已完全解析的自定义文件名或 `$DSH_HOME/settings.yaml` 文件名，并在文档缺失时以仅属主可访问的权限独占创建空文档，非文件提供方则保留基类的 `undefined`。仅限回环访问的 `settings.describe` 响应会在脱敏 namespace 视图旁只携带布尔型 `hasDocument` 能力。`ui-settings-general` 只在回环页面注册一条 `settings.action` 条目，只有元数据确认可准备好一份由提供方持有的本地文档后才显示，并调用无路径参数的 `settings.openDocument`；Host 会在文本文档交接前再次解析提供方路径（macOS 上使用 `open -t`，使任意 YAML 文件关联无法重定向这次操作；Linux 上使用 `xdg-open`；Windows 上使用 `Invoke-Item`）。通用 Workspace 路径仍保留现有的默认应用交接。浏览器既不推导 `$DSH_HOME`，也不会收到文件系统目标；远程页面不会为这项操作发起特权 settings 读取。
+**Host 识别并打开本地设置文档。** settings seam 暴露可选的 `documentPath` 提供方元数据和 `prepareDocument()` 操作；`settings-local` 返回已完全解析的自定义文件名或 `$DSH_HOME/settings.yaml` 文件名，并在文档缺失时以仅属主可访问的权限独占创建空文档，非文件提供方则保留基类的 `undefined`。仅限回环访问的 `settings.describe` 响应会在脱敏 namespace 视图旁只携带布尔型 `hasDocument` 能力。`ui-settings-general` 只在回环页面注册一条 `settings.action` 条目，只有元数据确认可准备好一份由提供方持有的本地文档后才显示，并调用无路径参数的 `settings.openDocument`；Host 会在文本文档交接前再次解析提供方路径（macOS 上使用 `open -t`，使任意 YAML 文件关联无法重定向这次操作；桌面 Linux 上使用 `xdg-open`；Windows 上使用 `Invoke-Item`；WSL 上先执行 `wslpath -w`，再使用同一 Windows 交接）。通用 Workspace 路径仍保留现有的默认应用交接。浏览器既不推导 `$DSH_HOME`，也不会收到文件系统目标；远程页面不会为这项操作发起特权 settings 读取。
 
 **llm seam 声明可配置性并公布拓扑。**`registerConfigurableProviders()` 是一个全有或全无、以 fiber 为作用域的目录，条目为 `{provider, displayName, settingsNs, settingsPath}`——这正是配置页要为一条可能尚不存在的路由打开正确设置子树时所需要的寻址；`listConfigurableProviders()` 在 wire 处理器里与存活路由合并，未声明的存活路由因此仍报告为激活。零负载的 `'llm/adapters-updated'` 事件从全部四个注册／注销提交点触发，listener 派发带异常隔离（INVARIANT 重抛），沿用 settings/commands 的先例。`llm-deepseek` 的路由重命名为 `deepseek-official`，因为 pi-ai catalog 名正言顺地拥有 `deepseek` 这个聚合器条目；依预发布立场，不设别名。
 
 **架在 schema 模型层之上的手写编辑器。**`dsh-client-schema-form` 把 wire 的 `toJSON()` 信封还原（rehydrate）为活的 schemastery 节点，用于校验、路径解析与不可变草稿编辑——但不做通用渲染：第一版交付了完整的 schema 驱动表单渲染器，得到的却是一个未加样式、把 schema 原样倾倒出来的页面（每个进阶字段都平铺到卡片上、原始字段名直接充当标签、`retryPolicy` 的「不支持」回退落在主流程里）。用户没有再加一套提示／分组系统，而是选择了手写方向，第二轮又把引用输入框整个移除：卡片的主字段是一个 **API 密钥**输入框，未配置密钥的整分节提供方会以其设置卡片的形式打开，收起的「自定义设置」折叠区承载按家族精选的额外字段（两个家族都有 `baseURL`，deepseek 有 `reasoningEffort`／pi-ai 有 `reasoning`，另有直接 DeepSeek 模型行的 `id`、`name` 和 `contextWindow`）。现有模型字段中不在可见集合内的部分会在数组编辑后保留；重试策略、超时及其他字段仍归 `settings.yaml` 所有。校验仍会在写入前运行还原出的 schema，适配器特有的检查则会拒绝序列化 schema 无法表达的目录不变量。卡片的颜色经 `--dsw-alias-*` 设计 token 解析；它此前引用的 `--border`／`--surface`／`--text-*` 在本应用中无人定义，于是渲染出的是它们的亮色模式回退值，在暗色主题下依旧保持亮色。模型目录采用 pi-ai 提供方表单引入的行形态：每个模型一个带边框的条目，ID 与显示名称落在行上，容量则收在该行自己的折叠区里，使两个编辑器呈现为同一套设计，而不是在双方都落地后各自分岔。每个字段都保留那个为其命名的带序号 `aria-label`。两项容量都是文本输入框，读取十进制的 `K`／`M` 后缀（`1M` 即 1000K，与容量的通行标注方式一致）并存储纯数值：字段持有焦点期间保留键入的文本，因为若每次按键都从解析出的数值重新推导该文本，`1000` 会在尚未输完时就被改写成 `1K`；无法解析的文本也会留在屏幕上，因此保存时的拒绝点名的是用户仍能看见的那一行。共用的类名承载的是本文件的 token 写法，而非那个分支的：`--dsw-alias-border-subtle`、`--dsw-alias-text-tertiary` 和 `--dsw-alias-text-primary` 均未声明，于是它们解析为各自回退槽位中的亮色模式字面值——正是本节此前迁离的那个缺陷。现在有一个样式测试会拒绝 token 表未声明的任何 `--dsw-*` 名称，因此下一个写出这类名称的编辑者会当场失败，而不是交付一个只有亮色的界面。
 
-**Models 页是一次三领域联接，应用语义与 seam 同形。**每一行是一个已配置的提供方；「新增」卡片的选择框是可配置提供方目录中剩余的休眠条目。路由存活状态仍用于就绪判定，并会使该联接失效，但页面不将其渲染为提供方状态，因为配置存在与运行时可用性是两个不同概念。密钥通道保持引用形态，却从不展示任何引用：键入的密钥经 `credentials.set` **只写**存入 profile 的 `apiKeyEnv` 之下，引用不存在时便派生 `<ROUTE>_API_KEY`（pi-ai profile 会记录该派生），因此 `settings.yaml` 从不携带密钥值。profile 的编辑和删除会针对脱敏后的用户分节，以按路径寻址的最小 `settings.mutate` 操作落地，绝不会点名页面未收到的机密。删除用户层提供方时，会先打开本地化的模型提供方确认对话框；取消操作、关闭按钮和遮罩均不会改动 profile，而破坏性确认会提交唯一一条 unset，并在其完成前阻止重复提交。DeepSeek 的模型列表是数组替换配置：继承而来的生效模型行会一直显示，直到第一次编辑将完整列表具化到用户层；重置则会取消设置该列表覆盖。
+**Models 页是一次三领域联接，应用语义与 seam 同形。**每一行是一个已配置的提供方；「新增」卡片的选择框是可配置提供方目录中剩余的休眠条目。路由存活状态仍用于就绪判定，并会使该联接失效，但页面不将其渲染为提供方状态，因为配置存在与运行时可用性是两个不同概念。密钥通道保持引用形态，却从不展示任何引用：键入的密钥经 `credentials.set` **只写**存入 profile 的 `apiKeyEnv` 之下，引用不存在时便派生 `<ROUTE>_API_KEY`（仅在输入密钥时，pi-ai profile 才会记录该派生），因此 `settings.yaml` 从不携带密钥值；留空 pi-ai 密钥会具化一个不带引用的 profile，并保留提供方原生认证。profile 的编辑和删除会针对脱敏后的用户分节，以按路径寻址的最小 `settings.mutate` 操作落地，绝不会点名页面未收到的机密。删除用户层提供方时，会先打开本地化确认对话框，其行操作、标题、说明和最终操作都会点名同一个提供方；确认后会先清除与派生目标精确匹配且已配置、可写的凭据，再删除 profile，自定义目标、环境目标和无法识别的目标则保持不变。两个阶段都具备幂等性，部分失败会留在对话框中供重试。DeepSeek 的模型列表是数组替换配置：继承而来的生效模型行会一直显示，直到第一次编辑将完整列表具化到用户层；重置则会取消设置该列表覆盖。部分提交与凭据所有权的理由记录在[提供方凭据生命周期 note](../bug-fix/2026-08-06-provider-credential-lifecycle.md)中。
 
 ## 曾考虑的替代方案
 
@@ -36,4 +36,4 @@ PR1 让 LLM（大语言模型）适配器配置在 seam 层面免重启，但唯
 
 ## 后果
 
-整条闭环以无密钥方式固定在浏览器测试通道（`apps/web/tests/models-settings.e2e.ts`）：「新增」卡片提供休眠的 pi-ai catalog，携键入的密钥添加 `minimax-cn` 会把只含引用的 profile 写入 `settings.yaml`、把密钥值存入 harness 家目录 `.env` 中派生的 `MINIMAX_CN_API_KEY` 之下、路由随拓扑帧注册为存活，「自定义设置」折叠区则把 `reasoning` 合并到引用旁边——全程零模型调用，「新增」卡片态、已配置态与删除确认态各有 ARIA golden，另有脚手架式的 `harnessHome`，测试绝不触碰真实的 `~/.dsh`（受测提供方是派生引用不可能与开发者已导出密钥相撞的那一个）。设置外壳场景会截获无路径参数的原生意图；seam、提供方、wire、React 与原生打开器测试分别固定了提供方缺失、自定义路径解析、缺失文件创建、仅属主权限、远程／不可用时隐藏、重复点击合并、本地化失败、macOS 文本编辑器分发，以及 Linux/Windows 桌面分发。删除场景证明：取消后 profile 保持原样，确认后会将其删除，而刻意保留的凭据依然存在。DeepSeek 首次使用 fixture 会把默认目录编辑为用户自有列表、持久化任意模型的 ID／名称／上下文窗口、移除活动模型行，并观察模型选择器的空选择回退。这次重命名在一次提交中触及 239 个文件（fixture（测试前置数据）、golden、文档、python），未保留兼容别名。替换渲染器只花了一次提交，且没有任何 wire 变更：应用语义、脱敏与目录联接从一开始就与渲染器无关。延后事项：每行的模型预览（选择器已能列出模型）、为从未声明可配置性的存活路由提供页面地址，以及显式删除提供方所保留的凭据。
+整条闭环以无密钥方式固定在浏览器测试通道（`apps/web/tests/models-settings.e2e.ts`）：「新增」卡片提供休眠的 pi-ai catalog，携键入的密钥添加 `minimax-cn` 会把只含引用的 profile 写入 `settings.yaml`、把密钥值存入 harness 家目录 `.env` 中派生的 `MINIMAX_CN_API_KEY` 之下、路由随拓扑帧注册为存活，「自定义设置」折叠区则把 `reasoning` 合并到引用旁边——全程零模型调用，「新增」卡片态、已配置态与已点名目标的删除确认态各有 ARIA golden，另有脚手架式的 `harnessHome`，测试绝不触碰真实的 `~/.dsh`（受测提供方是派生引用不可能与开发者已导出密钥相撞的那一个）。设置外壳场景会截获无路径参数的原生意图；seam、提供方、wire、React 与原生打开器测试分别固定了提供方缺失、自定义路径解析、缺失文件创建、仅属主权限、远程／不可用时隐藏、重复点击合并、本地化失败、macOS 文本编辑器分发，以及 Linux/Windows 桌面分发。删除场景证明，取消会保留 profile 和密钥，随后的确认会同时删除 profile 及其已识别的受管凭据。DeepSeek 首次使用 fixture 会把默认目录编辑为用户自有列表、持久化任意模型的 ID／名称／上下文窗口、移除活动模型行，并观察模型选择器的空选择回退。这次重命名在一次提交中触及 239 个文件（fixture（测试前置数据）、golden、文档、python），未保留兼容别名。替换渲染器只花了一次提交，且没有任何 wire 变更：应用语义、脱敏与目录联接从一开始就与渲染器无关。延后事项：每行的模型预览（选择器已能列出模型）和为从未声明可配置性的存活路由提供页面地址。

.agents/notes/implemented/architecture/2026-07-31-code-runtime-portable-identifier-seam.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/architecture/2026-07-31-code-runtime-portable-identifier-seam.md
+2026-07-31-code-runtime-portable-identifier-seam.md: 9e9df50235b3505458e3645e2c6ff6e9bd439183
+2026-07-31-code-runtime-portable-identifier-seam.zh.md: 31d2410ee4809d0693f2e7897e61c50163bb0758

.agents/notes/implemented/architecture/2026-07-31-code-runtime-portable-identifier-seam.md

@@ -0,0 +1,44 @@
+# Agent Note: the code-runtime seam owns portable-identifier exclusions
+
+Status: implemented
+
+English | [中文](2026-07-31-code-runtime-portable-identifier-seam.zh.md)
+
+## Problem
+
+The code-runtime seam promises that a binding-namespace list valid on one backend is valid on every backend, so a Code Mode consumer can hand the same bindings to any registered runtime without knowing its language. The first backend, `dsh-code-runtime-worker`, privately owned the identifier rules that enforce part of that promise: an `IDENTIFIER` regex that allowed the JS-only `$`, a `RESERVED_WORDS` set holding only ECMAScript keywords, and a `RESERVED_ERROR_PROPERTIES` set of three JS `Error` slots. Those rules described the worker's own language, not the seam's portability contract.
+
+A second backend written against a different language (CPython, arriving in a later PR of this stack) would either re-declare its own rules — letting `lambda` pass the worker and fail Python, or `$tools` pass the worker and fail every non-JS backend — or import the worker's, inverting the dependency so the seam's implementation reached into a sibling implementation. Neither keeps the portability promise real: it would hold only for the backend a caller happened to test against.
+
+## Decision
+
+The seam package (`@deepseek-ai/dsh-code-runtime`) exports the portable-identifier exclusion contract as four named constants, and every backend imports them rather than re-declaring:
+
+- `PORTABLE_RESERVED_WORDS` — the union of ECMAScript and Python reserved words. A namespace global or error-class name matching any is refused on all backends, so `lambda` is refused even though it is a legal JS parameter name. Adding a language widens this union, which is a deliberate breaking review of existing binding names.
+- `RESERVED_BINDING_GLOBALS` — globals some backend owns in the program's namespace: `console` (the worker's log capture), `__dsh_main__`/`__builtins__`/`__name__` (the Python bootstrap's wrapper and seeded module globals), and `__debug__` (not a seeded slot but a CPython compile-time constant that rejects assignment, so an injected global under that name is unreachable — the same portability split by a different mechanism). Refused everywhere so a namespace list cannot pick a name that works on one backend and collides on another.
+- `RESERVED_ERROR_MEMBERS` — error-member names every backend refuses: the JS `Error` slots (`name`, `message`, `stack`) and Python's exception-protocol members (`args`, `with_traceback`, `add_note`).
+- `DUNDER_MEMBER` — the dunder-form regex (`__x__`, non-empty middle), refused as an error member wholesale because several are constrained CPython descriptors whose exact set is an interpreter-version detail.
+
+The seam also narrows the portable identifier subset to `[A-Za-z_][A-Za-z0-9_]*` (documented on `CodeBindingNamespace.global` and `CodeBindingErrorClass`), dropping the JS-only `$`. The worker consumes the shared constants directly by their seam names — `PORTABLE_RESERVED_WORDS` for both binding-global and error-class names, `RESERVED_BINDING_GLOBALS` for backend-owned slots, `RESERVED_ERROR_MEMBERS` plus `DUNDER_MEMBER` for error members — with no local re-alias; its `IDENTIFIER` regex loses `$`.
+
+The constants live at the seam even though only one backend ships in this PR: the whole point is that the contract is language-agnostic and owned above any single language. A backend that violated it would be the bug, and the shared set is where a reviewer looks to see what "portable" means.
+
+## Scope
+
+This PR delivers only the seam extension and the worker's adoption of it. No Python backend, `py-types` renderer, or Code Mode language dispatch ships here — they are later PRs in the stack that depend on these exports. The seam README's worker-only wording is left unchanged for the same reason: linking to a `dsh-code-runtime-python` README that does not yet exist would break the dead-link gate.
+
+`RESERVED_BINDING_GLOBALS` currently encodes the not-yet-merged Python bootstrap's concrete design: it seeds exactly `__builtins__`/`__name__` and wraps the program under `__dsh_main__`. The Python-backend PR that seeds any additional module global (`__doc__`, `__loader__`, `__spec__`, `__file__`, `__package__`, …) MUST widen this set in the same change, exactly as adding a language widens `PORTABLE_RESERVED_WORDS` — a name the bootstrap seeds but the set omits is the portability split this contract exists to prevent.
+
+## Alternatives considered
+
+**Each backend declares its own exclusions.** Rejected: it makes the portability promise per-backend. A binding list the caller tested on the worker could be refused by Python, which is exactly the split the seam exists to prevent.
+
+**The Python backend imports the worker's constants.** Rejected: it inverts the dependency — the seam's implementations would reach into a sibling implementation for a contract neither owns. The contract belongs above both, at the seam.
+
+**Keep `$` in the portable identifier subset.** Rejected: `$` is JS-only spelling. Allowing it would let `$tools` pass the worker and fail every non-JS backend, breaking portability for a purely cosmetic gain.
+
+## Consequences
+
+Bought: one place — the seam package — defines what a portable binding name is, and every backend enforces the same contract by import. A namespace list valid on one backend is valid on all, verifiably, not by coincidence of which backend the caller tested.
+
+Cost: existing worker callers using a `$`-containing global now fail identifier validation. Under the pre-release stance this is a corrected foundation, not a compatibility break to shim. The worker's seam-misuse tests gain cases for `$tools`, Python exception members (`args`), dunders (`__dict__`), and a Python-owned global (`__dsh_main__`), proving the shared set is enforced from the worker side.

.agents/notes/implemented/architecture/2026-07-31-code-runtime-portable-identifier-seam.zh.md

@@ -0,0 +1,44 @@
+# Agent Note：code-runtime seam 拥有可移植标识符排除集
+
+Status: implemented
+
+[English](2026-07-31-code-runtime-portable-identifier-seam.md) | 中文
+
+## Problem
+
+code-runtime seam 承诺：在一个后端上有效的绑定命名空间列表，在每个后端上都有效，因此 Code Mode 消费方可以把同一组绑定交给任何已注册的运行时，而不必知道它的语言。首个后端 `dsh-code-runtime-worker` 私自拥有了执行这项承诺一部分的标识符规则：一个允许 JS 专有 `$` 的 `IDENTIFIER` 正则、一个只含 ECMAScript 关键字的 `RESERVED_WORDS` 集合，以及一个含三个 JS `Error` 槽位的 `RESERVED_ERROR_PROPERTIES` 集合。这些规则描述的是 worker 自身的语言，而非 seam 的可移植性契约。
+
+一个针对不同语言（CPython，将在本 stack 后续 PR 中到来）编写的第二后端，要么重新声明自己的规则——让 `lambda` 通过 worker 却在 Python 上失败，或让 `$tools` 通过 worker 却在每个非 JS 后端上失败——要么导入 worker 的规则，从而反转依赖，使 seam 的一个实现伸手进入另一个兄弟实现。二者都无法让可移植承诺成真：它只对调用方恰好测试过的那个后端成立。
+
+## Decision
+
+seam 包（`@deepseek-ai/dsh-code-runtime`）以四个具名常量导出可移植标识符排除契约，每个后端导入它们而非重新声明：
+
+- `PORTABLE_RESERVED_WORDS`——ECMAScript 与 Python 保留字的联集。任何命名空间 global 或 error-class 名称匹配其中之一，都在所有后端上被拒绝，因此 `lambda` 即便是合法的 JS 参数名也被拒绝。新增一门语言即扩宽此联集，这是对现有绑定名称的一次有意的破坏性复审。
+- `RESERVED_BINDING_GLOBALS`——某个后端在程序命名空间中拥有的 global：`console`（worker 的日志捕获）、`__dsh_main__`/`__builtins__`/`__name__`（Python bootstrap 的包装器与预置模块 global），以及 `__debug__`（不是 seed 的槽位，而是 CPython 编译期常量，赋值会被拒，故以该名注入的 global 不可达——同一种可移植性分裂，只是机制不同）。在所有后端上被拒绝，使命名空间列表无法选到一个在某后端能用、在另一后端冲突的名称。
+- `RESERVED_ERROR_MEMBERS`——每个后端都拒绝的 error-member 名称：JS `Error` 槽位（`name`、`message`、`stack`）与 Python 异常协议成员（`args`、`with_traceback`、`add_note`）。
+- `DUNDER_MEMBER`——dunder 形式正则（`__x__`，非空中缀），作为 error member 被整体拒绝，因为其中若干是受约束的 CPython 描述符，其确切集合是解释器版本细节。
+
+seam 同时把可移植标识符子集收窄为 `[A-Za-z_][A-Za-z0-9_]*`（记录在 `CodeBindingNamespace.global` 与 `CodeBindingErrorClass` 上），去掉 JS 专有的 `$`。worker 直接以 seam 名消费这些共享常量——binding-global 与 error-class 名称用 `PORTABLE_RESERVED_WORDS`、后端拥有槽位用 `RESERVED_BINDING_GLOBALS`、error member 用 `RESERVED_ERROR_MEMBERS` 加 `DUNDER_MEMBER`——不再本地起别名；其 `IDENTIFIER` 正则去掉 `$`。
+
+尽管本 PR 只交付一个后端，这些常量仍置于 seam：要点正是该契约与语言无关，且拥有权在任何单一语言之上。违反它的后端才是 bug，而共享集合正是复审者查看"可移植"含义的地方。
+
+## Scope
+
+本 PR 只交付 seam 扩展与 worker 对它的采用。这里不交付任何 Python 后端、`py-types` 渲染器或 Code Mode 的语言分发——它们是本 stack 中依赖这些导出的后续 PR。seam README 中仅描述 worker 的措辞保持不变，理由相同：链接到一个尚不存在的 `dsh-code-runtime-python` README 会破坏死链 gate。
+
+`RESERVED_BINDING_GLOBALS` 当前编码了尚未合并的 Python bootstrap 的具体设计：它恰好 seed `__builtins__`/`__name__`，并把程序包装在 `__dsh_main__` 之下。任何 seed 额外模块 global（`__doc__`、`__loader__`、`__spec__`、`__file__`、`__package__` 等）的 Python 后端 PR 必须在同一改动中扩宽此集合，正如新增一门语言即扩宽 `PORTABLE_RESERVED_WORDS`——bootstrap 会 seed 却不在集合中的名称，正是本契约要防止的可移植性分裂。
+
+## Alternatives considered
+
+**每个后端声明自己的排除集。** 拒绝：这让可移植承诺变成逐后端成立。调用方在 worker 上测过的绑定列表可能被 Python 拒绝，而这正是 seam 存在要防止的分裂。
+
+**Python 后端导入 worker 的常量。** 拒绝：这反转依赖——seam 的实现会为一个二者都不拥有的契约伸手进入兄弟实现。契约属于二者之上，即 seam。
+
+**在可移植标识符子集中保留 `$`。** 拒绝：`$` 是 JS 专有拼写。允许它会让 `$tools` 通过 worker 却在每个非 JS 后端上失败，为纯粹表面的好处破坏可移植性。
+
+## Consequences
+
+获得：一个地方——seam 包——定义什么是可移植绑定名称，每个后端通过导入执行同一契约。在一个后端上有效的命名空间列表在所有后端上都有效，这是可验证的，而非取决于调用方测试了哪个后端的巧合。
+
+代价：现有使用含 `$` global 的 worker 调用方现在会在标识符校验时失败。在预发布立场下这是一次被纠正的地基，而非需要 shim 的兼容性破坏。worker 的 seam-misuse 测试新增了 `$tools`、Python 异常成员（`args`）、dunder（`__dict__`）与一个 Python 拥有的 global（`__dsh_main__`）等用例，从 worker 侧证明共享集合被执行。

.agents/notes/implemented/architecture/2026-08-05-profile-plugin-bundles.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/architecture/2026-08-05-profile-plugin-bundles.md
+2026-08-05-profile-plugin-bundles.md: 11a8ac3d4005371ca9596ba237aaf42a8e770dee
+2026-08-05-profile-plugin-bundles.zh.md: 0e9ebf657ccb9d05967d90a935b356acf287a24c

.agents/notes/implemented/architecture/2026-08-05-profile-plugin-bundles.md

@@ -0,0 +1,33 @@
+# Agent Note: Profile plugin bundles replace the fixed surface overlays
+
+Status: implemented
+
+English | [中文](2026-08-05-profile-plugin-bundles.zh.md)
+
+## Problem
+
+The `dsh` launcher hardcoded its compositions: `base.cordis.yml` + `web.cordis.yml` shipped inside `apps/cli`, three bespoke entry modes (`--config`, `web`, `-p`) each with its own layer stack, and a single global personal overlay (`$DSH_HOME/config.yaml`). There was no way to install an out-of-tree plugin (a TUI, a provider pack) into a shipped surface without editing the repository, and no place where a third-party package could contribute a default composition.
+
+## Decision
+
+Everything becomes a **profile**: a directory `$DSH_HOME/profiles/<name>` with a `package.json` (pnpm-managed out-of-tree plugin `dependencies` plus the profile manifest `dsh.profile` with its ordered `bundles` layer list) and a user `cordis.patch.yml`. A **bundle** is an npm package declaring `"dsh": { "bundle": { "patch": "./cordis.patch.yml" } }`; the two manifest kinds live under distinct `dsh.profile` / `dsh.bundle` keys so a package.json states which role it plays. The tree composes over an empty root by applying each bundle's patch in `dsh.profile.bundles` order, then the user layer, then `--patch` overlays, then flag patches — one `applyEntryPatches` call, identical for boot, flag derivation, and `--dump-config`.
+
+The shipped compositions became bundles: `@deepseek-ai/dsh-base` (the former base rows as one insert), `@deepseek-ai/dsh-web-app` (the former web overlay plus a runtime glue plugin that owns what used to be launcher code — frontend-dist resolution, the web-surface prompt section, bash runtime variables, the URL line), and `@deepseek-ai/dsh-headless` (a one-shot runner plugin over base + web-app). `dsh web` stays as an alias for `--profile web` carrying the Web flag family; `dsh --profile headless "task"` replaces `-p`; `dsh --config` is removed (its uses migrate to `--patch`). `dsh plugin --profile <name> <args...>` is a thin pnpm forwarder that initializes the profile and reconciles `dsh.profile.bundles` after `add`/`remove` (a bundle-less package warns and stays a plain dependency).
+
+Resolution is two-anchored by construction: `dsh.profile.bundles` names resolve from the dsh installation first, then the profile directory — so in-box bundles always come from the same installation as the running `dsh` and pnpm never manages them — while bare plugin names in patch rows resolve through the profile directory's Node parent-walk into the maintained flat fallback `$DSH_HOME/profiles/node_modules` (one symlink per package the installation's app and bundles depend on, healed on every launch).
+
+Two supporting refactors: the webserver's built-in static dist serving became the single-owner **fallback seat** (`registerFallback`/`applyIndexTaps`), with the SPA server extracted to `@deepseek-ai/dsh-frontend-static` so the web bundle owns its dist as composition, not launcher code; and the personal-overlay machinery of the [dsh CLI personal-config decision](../feature/2026-07-20-dsh-cli-personal-config.md) (`loadPersonalPatches`, `$DSH_HOME/config.yaml`) was retargeted to the per-profile and home-level `cordis.patch.yml` layers (`loadOptionalPatches`, `watchUserPatches` taking a filename), superseding that note's entry modes and file location while keeping its Harness-home root, patch semantics, and fail-loud parsing.
+
+## Alternatives considered
+
+- **Dependency-scan plus partial `patchOrder`** (the original sketch): scanning `dependencies` for bundles and ordering unlisted ones alphabetically has two sources of truth and an implicit tie-break; one explicit ordered `dsh.profile.bundles` list is smaller and fully deterministic. A raw `pnpm add` inside the profile installs a library without activating any patch — explicit, no spooky scan.
+- **`link:` entries for in-box bundles**: pnpm cannot version, install, or update a `link:` into the installation, it embeds a machine path in a user file, and it breaks when the installation moves. The two-anchor resolution plus healed symlink fallback gives the same guarantee ("bundles come from the installation") without ceremony.
+- **A pre-boot `context` module in the bundle manifest** for boot-time values (dist path, flag facts): rejected in favor of pure plugins — the glue is ordinary rows the launcher patches, so the composition stays fully dumpable and the manifest stays data-only. The launcher-owned `ctx.headlessIo` seam is the one host-provided slot, and it is provided in `boot()`'s `prepare` hook, before any config-tree entry mounts.
+- **Transitive bundle auto-application**: only direct `dsh.profile.bundles` entries contribute layers; a meta-bundle wanting to re-export another bundle's patch must do so explicitly in its own patch file.
+
+## Consequences
+
+- New composition surfaces (a TUI, provider packs) ship as ordinary npm packages installable per profile; the repository no longer needs a row for every deployment shape.
+- `apps/cli` shrank to argv parsing, profile machinery consumption, and the pnpm forwarder; `AppCLIEntry` and the per-surface boot paths are gone.
+- The keyless web e2e scaffold boots the same bundle layers over the same empty-root shape as production, including the profiles module fallback, so composition drift between test and product fails loudly.
+- Backends reject nothing old on disk (pre-release stance): `$DSH_HOME/config.yaml` is simply no longer read.

.agents/notes/implemented/architecture/2026-08-05-profile-plugin-bundles.zh.md

@@ -0,0 +1,33 @@
+# Agent Note: profile 插件组合包取代固定的表层 overlay
+
+Status: implemented
+
+[English](2026-08-05-profile-plugin-bundles.md) | 中文
+
+## Problem
+
+`dsh` 启动器硬编码了自己的组合：`base.cordis.yml` + `web.cordis.yml` 随 `apps/cli` 一起交付，三种各自定制的入口模式（`--config`、`web`、`-p`）各带一套层栈，外加一个全局的个人 overlay（`$DSH_HOME/config.yaml`）。想把树外插件（一个 TUI、一个提供方扩展包）装进已交付的表层，只能修改仓库；第三方包也没有任何位置可以贡献默认组合。
+
+## Decision
+
+一切都变成 **profile**：即目录 `$DSH_HOME/profiles/<name>`，其中包含一个 `package.json`（pnpm 管理的树外插件 `dependencies`，加上 profile manifest `dsh.profile` 及其有序的 `bundles` 层列表）和一份用户 `cordis.patch.yml`。**组合包**（bundle）是声明了 `"dsh": { "bundle": { "patch": "./cordis.patch.yml" } }` 的 npm 包；两种 manifest 分别位于互不相同的 `dsh.profile` / `dsh.bundle` 键下，因此一份 package.json 能说明自己扮演哪种角色。配置树在空的根之上组合：按 `dsh.profile.bundles` 顺序应用每个组合包的 patch，然后是用户层，然后是 `--patch` overlay，最后是 flag patch——全部收敛为一次 `applyEntryPatches` 调用，启动、flag 派生与 `--dump-config` 使用完全相同的路径。
+
+已交付的组合改造成了组合包：`@deepseek-ai/dsh-base`（原有基础行合并为一次插入）、`@deepseek-ai/dsh-web-app`（原 web overlay，外加一个接管原启动器代码的运行时粘合插件——前端 dist 解析、web 表层提示词段落、bash 运行时变量、URL 行）、`@deepseek-ai/dsh-headless`（叠加在 base + web-app 之上的一次性 runner 插件）。`dsh web` 保留为携带 Web flag 家族的 `--profile web` 别名；`dsh --profile headless "task"` 取代 `-p`；`dsh --config` 被移除（其用途迁移到 `--patch`）。`dsh plugin --profile <name> <args...>` 是一层薄薄的 pnpm 转发器，负责初始化 profile，并在 `add`／`remove` 后调和 `dsh.profile.bundles`（没有组合包声明的包会给出警告，保持为普通依赖）。
+
+解析在构造上就是双锚点的：`dsh.profile.bundles` 中的名称先从 dsh 安装目录解析，再从 profile 目录解析——因此内置组合包始终来自与运行中 `dsh` 相同的安装，pnpm 从不管理它们——而 patch 行中的裸插件名称经 profile 目录的 Node 父目录逐级查找，落到受维护的扁平回退目录 `$DSH_HOME/profiles/node_modules`（安装目录的应用与各组合包所依赖的每个包各一个符号链接，每次启动时修复）。
+
+两项配套重构：webserver 内置的静态 dist 服务改为单一所有者的**回退席位**（`registerFallback`／`applyIndexTaps`），SPA 服务器提取到 `@deepseek-ai/dsh-frontend-static`，使 web 组合包以组合的方式持有自己的 dist，而不是靠启动器代码；[dsh CLI 个人配置决策](../feature/2026-07-20-dsh-cli-personal-config.md)的个人 overlay 机制（`loadPersonalPatches`、`$DSH_HOME/config.yaml`）改为面向逐 profile 与 home 级的 `cordis.patch.yml` 层（`loadOptionalPatches`、接受文件名的 `watchUserPatches`），取代该笔记的各入口模式与文件位置，同时保留其 Harness home 根目录、patch 语义与大声失败的解析。
+
+## Alternatives considered
+
+- **依赖扫描加部分 `patchOrder`**（最初的草案）：扫描 `dependencies` 找出组合包、未列出者按字母序排列，会产生两个真源和一条隐式决胜规则；一份显式有序的 `dsh.profile.bundles` 列表更小、完全确定。在 profile 内直接 `pnpm add` 只会安装一个库，不激活任何 patch——行为显式，没有暗中扫描。
+- **内置组合包使用 `link:` 条目**：pnpm 无法对指向安装目录的 `link:` 做版本管理、安装或更新，它会把机器路径嵌进用户文件，并且在安装目录移动后失效。双锚点解析加上每次启动修复的符号链接回退提供了同样的保证（「组合包来自安装目录」），且没有这些繁文缛节。
+- **在组合包 manifest（元数据清单）中放一个启动前 `context` 模块**承载启动期取值（dist 路径、flag 事实）：否决，改用纯插件——粘合逻辑就是启动器 patch 的普通配置行，因此组合始终可完整 dump，manifest 保持纯数据。启动器持有的 `ctx.headlessIo` seam 是唯一由宿主提供的 slot，且在任何配置树条目挂载之前，于 `boot()` 的 `prepare` 钩子中提供。
+- **组合包的传递式自动应用**：只有直接列在 `dsh.profile.bundles` 中的条目才贡献层；想重新导出另一个组合包 patch 的元组合包，必须在自己的 patch 文件中显式完成。
+
+## Consequences
+
+- 新的组合表层（TUI、提供方扩展包）以普通 npm 包形式交付，可按 profile 安装；仓库不再需要为每种部署形态各留一行。
+- `apps/cli` 收缩为 argv 解析、profile 机制的消费方和 pnpm 转发器；`AppCLIEntry` 与各表层专属的启动路径全部移除。
+- 无密钥 web e2e 脚手架以与生产相同的空根形态启动相同的组合包层，包括 profiles 模块回退，因此测试与产品之间的组合漂移会大声失败。
+- 后端不拒绝磁盘上的任何旧格式（发布前姿态）：`$DSH_HOME/config.yaml` 只是不再被读取。

.agents/notes/implemented/architecture/2026-08-06-subagent-list-identity-projection.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/architecture/2026-08-06-subagent-list-identity-projection.md
+2026-08-06-subagent-list-identity-projection.md: ba023d04805ff3335c8f243510aa8ddc15fe13d6
+2026-08-06-subagent-list-identity-projection.zh.md: 368a70f5b3e5a27e4e1c648e8819476d40a57709

.agents/notes/implemented/architecture/2026-08-06-subagent-list-identity-projection.md

@@ -0,0 +1,184 @@
+# Agent Note: Subagent list identity via the projection unit
+
+Status: implemented
+
+English | [中文](2026-08-06-subagent-list-identity-projection.zh.md)
+
+## Problem
+
+Before the rewrite, `SubagentService.listChildren` ran two full-log materializations — `listEvents` plus `readEvent` — on every listing for each direct child with `header.origin === 'subagent'`, each materialization accompanied by a full-log structuredClone, all to fold two fields, mode and label, out of the descriptor event. The descriptor's position in the log is not fixed — the fork prefix is arbitrarily long, and zstd-compressed frames carry no seq index — so there is no shortcut to locating it; this path had no cache whatsoever, and its cost amplifies with transcript length × child count × listing frequency. It also dragged session-query in as a hard dependency of listing: in a deployment without a query backend, `list_agents` rejects wholesale with `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE`, even though enumeration needs nothing but header facts.
+
+The same root cause has a second symptom: on every Agent-bound RPC's owner check, the host-side `hasSubagentDescriptor()` scans the target session's own suffix, even though `SessionHeader.origin` already answers the vast majority of the same question.
+
+The root cause is that the [durable-subagent-catalog decision](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md) made the descriptor event (`subagent/descriptor`) the catalog's sole durable authority yet paired descriptor reads with no cache layer, and explicitly accepted the per-child double read as the "no-index correctness baseline". [Web subagent conversations](../feature/2026-07-27-web-subagent-conversations.md) (#1569) already put "is this a subagent" into the header (`SessionHeader.origin`), so identity determination no longer reads the log; mode and label still had to be scanned.
+
+## Decision
+
+mode and label are folded by the new `subagent` projection unit (pure identity, two arms), and the unit is the sole authority over the fold rules; `listChildren` no longer depends on session-query — enumeration is a subagent-owned live-preferred merge, and value retrieval walks a three-rung compute-and-discard ladder: a live child synchronously reads the registry's existing watermark cache (zero log reads); a cold child first asks the optional `sessionProjectionCache` checkpoint, and a served identity that passes the seq gate is final; otherwise it pays one full `persistence.inspect` read plus one `registry.restore` fold. No index, no cache of its own, no write-back.
+
+There are three families of escape from the per-child scan: promote mode/label into the header (the write path pays); build a durable derivation for the projection (a checkpoint ladder, or values landed during query-index rebuild with read-side reconciliation); or compute at read time (live from the watermark cache, cold from one full read). This note takes the third. "Values landed with the query index" was once this note's settled direction and was under construction for a time, then retired wholesale: query infrastructure was forced to learn domain vocabulary while the sole consumer is satisfied by read-time computation — the live child's zero reads come for free from session-projection's existing watermark cache, and the cold child's single full read is explicitly accepted as compute-and-discard. The first two routes and the retirement rationale are detailed under Alternatives considered.
+
+Key points:
+
+- **The subagent list does not depend on session-query**: enumeration is completed by a subagent-owned live-preferred merge, and mode/label is retrieved through `ctx.sessionProjections`; deployments without a query backend list as usual.
+- **Value retrieval is a three-rung compute-and-discard ladder**: a live child reads `sessionProjections.snapshot()` (the registry's existing watermark cache, zero log reads); a cold child first reads the optional `sessionProjectionCache.cachedSnapshot(header)`, using the value directly when a non-null `subagent` identity passing the seq gate (`seq >= seedLength ?? 0`) is among its values; otherwise it pays one full `persistence.inspect` read plus one `registry.restore({}, events, 0)` fold; beyond that, absent is absent — no cache of its own, no write-back, no index.
+- **The `subagent` projection unit is the sole authority over the fold rules**: the live snapshot, the cold restore, and GUI history's detached fold all compute through the registry; no second copy of descriptor-interpretation logic exists.
+- **The header, the descriptor (v2), session-persistence, session-projection(-cache), and session-query(-sqlite) are all untouched**; pre-existing data acquires exact values through one `inspect` computation the first time it is listed — no degraded unknown state, no migration.
+
+Relationship to existing notes:
+
+- This note supersedes two designs on the list read path in [durable-subagent-catalog](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md): enumeration through `sessionQuery.traceSession`, and per-child descriptor-event reads (the `listEvents`-plus-exact-`readEvent` double read with in-place diagnostic classification). The diagnostic row semantics is retained, with classification now derived by the list from projection-value absence and activity; the descriptor event remains the sole durable authority for mode/label and the fold input, and the resume authorization and Activation contracts are untouched. This is partial supersession; the two notes stay cross-linked.
+- The [session-projection RFC](../../proposed/architecture/2026-07-27-session-projection-and-command-log.md)'s registry contract (`ProjectionDefinition`, `snapshot`, `restore`) is untouched; this note only adds one registration to it — the `subagent` identity unit — and becomes another consumer instance of the two existing reads, snapshot (live) and restore (cold) — GUI history's cold read is already the same shape. The fold rules are registered with the registry exactly once; every consuming surface computes through the registry, and no second copy of the fold logic exists.
+
+### `subagent` projection unit
+
+It hangs beside the existing `subagentTiming` ([projection.ts](../../../../packages/subagent/subagent/src/projection.ts), [projection-types.ts](../../../../packages/subagent/subagent/src/projection-types.ts)), under key `subagent`:
+
+```ts ignore-check
+export type SubagentIdentityProjection =
+  | { mode: 'one-shot'; label?: string; seq: number }
+  | { mode: 'continuable'; label: string; seq: number }
+
+declare module '@deepseek-ai/dsh-session-projection/types' {
+  interface SessionProjectionMap {
+    subagent: SubagentIdentityProjection | null
+  }
+}
+```
+
+- The projection is pure identity, and **the projection system has no failure channel**: a unit never throws; a corrupt payload or an unrecognized version folds exactly like a log with no descriptor at all — the result is a **serializable null sentinel**: the map entry is `SubagentIdentityProjection | null`, non-optional, never undefined or an absent key. The reason: the registry's onChanged push goes through JSON serialization, where an undefined field is dropped by stringify, the client's frame validation rejects the frame, and a consumer's stored old identity would never update; null passes frames intact, and consumers replace the old identity with the sentinel. The judging discipline: consuming surfaces treat null and undefined (which only a JSON boundary dropping the key can produce) alike as no value. How "computed to nothing" is presented is the consumer's own business (see the `listChildren` four-state mapping below).
+- Label strength is decided by the descriptor schema: a continuable's label is mandatory at parse, a one-shot's was always optional; the mode/label discriminant matches the child row's strong contract below exactly (the row carries no `seq` — it is the projection's internal own-suffix proof).
+- The identity carries `seq`: the seq of the `subagent/descriptor` event it was folded from, mandatory on both arms and absent on the null sentinel — `seq >= header.seedLength ?? 0` proves the identity was folded from the child's own suffix rather than a fork seed's replayed ancestor descriptor. The state gaining `seq` bumps the unit's `stateVersion` to 2, and existing checkpoint rows are invalidated by version mismatch per the registry contract, falling to the authoritative refold.
+- Fold rule: `subagent/descriptor` is last-wins, under the same descriptor-reset discipline as `subagentTiming` — ancestor descriptors in the fork prefix are overridden by the session's own descriptor. A corrupt or unrecognized-version payload is last-wins all the same: it resets to the null sentinel rather than keeping the prior identity, so a fork of a healthy ancestor does not inherit an identity its own descriptor cannot stand up.
+
+### Enumeration: subagent-owned live-preferred merge
+
+`listChildren`'s ([list-children.ts](../../../../packages/subagent/subagent/src/list-children.ts)) enumeration goes through no query service: the two sources `ctx.sessions.list()` and `ctx.get('sessionPersistence')?.list()` merge by id, with a live record overriding the same-id persisted record wholesale and no header consistency check. Everything enumeration needs is header facts:
+
+- Filtering: `header.origin === 'subagent' && header.parentSession === parentSessionId`.
+- `hasChildren`: the same merged material, looked at one level down — a direct descendant exists with `origin === 'subagent'` whose `parentSession` is that child.
+- `activity`: a live record is `running`; one present only in persistence is `inactive`.
+- Ordering: `createdAt` ascending, then child id ascending (matching the old contract).
+- **Absent persistence degrades to live-only enumeration, not an error**: in a deployment without persistence, a cold child could not be resumed anyway, and listing live children remains meaningful. (Contrast: the old implementation rejected wholesale when sessionQuery was missing.)
+- A persistence listing failure fails the whole enumeration; per-child isolation applies only to the per-child cold reads.
+
+### Value retrieval: the three-rung compute-and-discard ladder
+
+For each enumerated child, mode/label retrieval walks a three-rung ladder — compute-and-discard, no cache of its own, no write-back (the third rung is the same shape as apiproxy `session.history`'s cold read):
+
+| Rung | Read | Cost |
+| --- | --- | --- |
+| 1: live child | `ctx.sessionProjections.snapshot(session).values.subagent` | Zero log reads — the registry's existing watermark cache, synchronous retrieval |
+| 2: cold child, cache hit | The optional `sessionProjectionCache.cachedSnapshot(header)`, used directly only when a non-null `subagent` identity satisfies `identity.seq >= header.seedLength ?? 0` — an own descriptor is immutable once appended, and the seq gate proves the value was folded from the child's own suffix, regardless of the row's watermark | Zero log reads |
+| 3: cold child, fallback | One full `persistence.inspect(id)` read + `registry.restore({}, events, 0).snapshot.values.subagent` | One full read computed per listing |
+
+- Error contract: an unmounted `ctx.sessionProjections` is a configuration error; `listChildren` checks unconditionally before enumerating and fails loudly with `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE` — a deployment with zero children fails just as deterministically, so an empty listing cannot mask the misconfiguration. The session store gets the same posture: an absent `ctx.get('sessions')` (a strict global read, never the caller-scope-bound property proxy) fails with `SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE`. The two codes map differently on the wire: apiproxy gives only `PROJECTIONS_UNAVAILABLE` a dedicated wire face, and `SESSION_STORE_UNAVAILABLE` goes through the generic internal fallback — the apiproxy composition injects `sessions` itself, so that error is unreachable in its deployment, and a dedicated mapping would violate the need principle. `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` is deleted along with the session-query dependency.
+- The cache is a purely optional acceleration layer: an absent service is skipped on a null check — no error code, no part in configuration validation (in contrast to `sessionProjections`' loud contract). Anything the second rung throws (including a poisoned unit row in the cache detonating `viewCheckpoint`) silently falls to the third rung — the cache is derived data, so its faults never produce a `corrupt` verdict; the final judgment belongs to the authoritative refold. A row whose checkpoint cut predates the descriptor naturally lacks the `subagent` key and falls through automatically, with no special-casing; a null sentinel in the row does not count either — it falls to the third rung for the authoritative refold's verdict. A count/interval checkpoint inside the creation window can land a fork seed's replayed ancestor identity in the row — the ancestor's seq falls inside the seed range, the seq gate rejects it, and it likewise falls to the third rung's verdict.
+- Per-child isolation: a single child's failed cold full read only turns that row into an `unavailable` diagnostic, naturally retried on the next listing, without affecting siblings (see the four-state mapping).
+- The cold path's lifecycle witness: preparation's result must still point at the lifecycle that was enumerated — the witness field set is the same seven fields as the old SOURCE_CONFLICT check (version, id, createdAt, cwd, parentSession, seedLength, delegationDepth); a session deleted and republished under the same id degrades to a `corrupt` row in the old parent's catalog, leaking nothing of the new owner's child.
+- Cold-read concurrency is bounded by the constant 4 — it constrains a read-only scan of local media, not deployment behavior; when a networked persistence backend appears, it is promoted to a validated `Config` field.
+- The cold-read cost, recorded honestly: only with the cache unmounted or missed does a cold child pay one full read per listing, at a cost proportional to its transcript size; the settled stance is compute-and-discard, and no cache of its own is built. The full read goes through `inspect()` into the [Session preparation](2026-08-05-session-preparation.md) cold read, so short-term repeated reads of the same id can hit its LRU for reuse, but listing does not depend on this. A live child reads zero log throughout.
+- Cancellation: the caller's signal is checked before and after each persistence read, and a read that settles only after abort is rejected, normalized to the stable error code `CANCELLED`.
+
+### Authority model
+
+- The session log is the sole authority; this design adds no derived persistence of any kind — no index values, no checkpoints of its own, no in-process memo; the `sessionProjectionCache` checkpoint the second rung reads is an existing composition item's derived data, which this design only reads and never writes. Values are computed on read and discarded, and a value's freshness is exactly the live state or persisted revision at the moment of the read (an own descriptor is immutable once appended — a cached identity past the seq gate has no staleness problem; the gate guards against seed-replayed ancestor identities).
+- The Session and persistence write paths are entirely unaware of listing and projection consumption: no event-listener write-back, no fold-on-write.
+- Enumeration and value retrieval constitute no second authorization source and make no unpublished child visible — the two sources see only published live records and durably written persisted records, consistent with the rule the durable-subagent-catalog note laid down for derived read surfaces.
+
+### `listChildren` row shape and consuming surfaces
+
+The `SubagentListEntry` **data structure is identical to before the rewrite** — the child and diagnostic arms, the `kind` discriminant, the three-valued `reason`, and the child arm's strong mode/label contract are all retained; the only change is the diagnostics' information source: the projection system has no failure channel, so diagnostics are derived by the list from projection-value absence and activity, and the list itself parses zero events. The "no value means await the hard read" rule guarantees the ladder always computes mode/label for healthy data.
+
+```ts ignore-check
+export type SubagentListEntry =
+  | ({
+    readonly kind: 'child'
+    readonly id: SessionId
+    readonly activity: 'running' | 'inactive'
+    readonly hasChildren: boolean
+  } & (
+    | { readonly mode: 'one-shot'; readonly label?: string }
+    | { readonly mode: 'continuable'; readonly label: string }
+  ))
+  | {
+    readonly kind: 'diagnostic'
+    readonly id: SessionId
+    readonly reason: 'corrupt' | 'unsupported' | 'unavailable'
+  }
+```
+
+For each enumerated child, the ladder's result maps to a row through four states:
+
+| Ladder result | Row |
+| --- | --- |
+| Snapshot carries a non-null `subagent` identity | child row |
+| Snapshot present, `subagent` null sentinel or key absent, and the child is **inactive** | diagnostic row, reason `corrupt` (settled debris: a missing, corrupt, or unrecognized-version descriptor, no longer subdivided) |
+| Snapshot present, `subagent` null sentinel or key absent, and the child is **running** | no row (creation window: the descriptor is not yet appended — the same window the old implementation omitted) |
+| The cold full read fails | diagnostic row, reason `unavailable` |
+
+- `unsupported` is no longer produced: the type and the wire enum retain the member under "data structures stay as they are", and this note records it as no longer produced.
+- Descriptor-less settled debris moves from the old implementation's omit into the `corrupt` diagnostic — damaged, dead child sessions in the corpus are visible rather than silently vanishing, which is exactly the original motivation for keeping diagnostics.
+- Any registered unit whose fold/schema throws on this child's log is likewise contained as that child's diagnostic row, reason `corrupt` — a deterministic data fault, aligned with the old implementation's `SESSION_QUERY_CORRUPT_SESSION`→`corrupt` mapping semantics; live and cold are treated alike, isolation is per-child, and siblings and the listing itself are unaffected. It is orthogonal to "value absent + running → omit": the creation window means "no data yet", a fold throw means "the data is bad" — a poisoned running child also gets a `corrupt` row rather than an omit.
+
+Known boundary deviations (deliberately accepted, recorded with this note):
+
+- A fork child that died in its publication window, with an ancestor descriptor in its seed, gets the ancestor identity from last-wins and wrongly surfaces as a child row; resume still fails against the own-suffix fold authority (`NOT_RESUMABLE`). The old implementation omitted it via `seedLength` filtering; the projection unit cannot see the header, and this debris-grade deviation is accepted (`subagentTiming` has the same kind of pre-existing exposure).
+- Multiple descriptors in the own suffix: the old implementation judged corrupt; last-wins now takes the final one (the provider contract guarantees exactly one anyway).
+- A live/persisted header conflict: the old implementation made it per-child corrupt; enumeration now prefers live with no consistency check, the conflict goes unnoticed, and the live record forms the row.
+- A source-read failure on damaged storage (e.g. a bad surface rejected by the cold full read): the old implementation mapped it to per-child `corrupt`; it is now uniformly an `unavailable` row (the read side cannot tell the causes apart).
+- An unknown parent: the old implementation threw not-found through session-query ('parent session … was not found'); the subagent-owned merge now yields an empty subset for a nonexistent parent, enumeration returns an empty list, and later operations on the wire land as child-level subagent-not-found — a silent change of semantics and wording, recorded as explicitly accepted.
+- Rung 2's later-event window: a cache row lands right after the first own descriptor, the log then appends a second own descriptor (or a malformed payload setting the null sentinel), and the process crashes before the next checkpoint — from then on a cold listing's rung 2, admitted by the seq≥seedLength gate, keeps serving the row's old identity (the first own descriptor's value), diverging from the authoritative refold (last-wins, the second), and a rung-2 hit triggers no refold, so nothing notices. Three boundaries: ① the precondition is a second own descriptor on the same child, violating the establishing provider's append-exactly-once contract — corruption-class data, same family and source as the multi-descriptor deviation; ② it takes both "corruption + a crash missing every checkpoint (the two mandatory points, turn/end and disposal, and the count/interval throttle points all unmet)" at once; ③ a healthy child (exactly one own descriptor) is unaffected — what the seq gate admits is precisely the only true identity. Self-healing: any live run of that child (the turn/end mandatory checkpoint) or any moment that triggers cache.write overwrites the whole row with a fresh fold (whole-record replace), and rung 2 serves correctly from then on; the authoritative paths (the rung-3 refold, the live snapshot, the resume fold) are correct from the start, and the divergence exists only in listing reads while the child stays cold and the row is never rewritten. The mechanical fixes were not taken: gate reconciliation would need the log-end seq, unavailable to a zero-read cold path; a cache row carrying the revision is an opaque token, incomparable and a cross-domain schema change — filed as accepted under the "the cache is never authoritative" doctrine.
+
+Consuming surfaces: diagnostic handling across wire, tool, and GUI **stays entirely as it was, zero changes** (the `list_agents` description and output schema are untouched; the plugin only narrows its load requirement — `sessionQuery` dropped from inject). The only behavioral changes are in apiproxy: on the route segment, the `hasSubagentDescriptor()` scan is deleted and `hasSubagentOwner` looks only at `header.origin` — pre-#1569 data without `origin` is no longer recognized as a subagent owner; it never entered the catalog anyway, and the pre-release stance accepts this; and `subagents.history` is aligned with `session.history`'s source — a live child served from in-memory events and the registry's watermark snapshot, a cold child from `inspectServable` reading persistence directly with a detached fold, no query service involved, the SESSION_QUERY_* error arms retired with it, and the wire shape unchanged (the `history` JSDoc wording becomes the live in-memory snapshot / cold persisted log dual arm).
+
+### Change footprint
+
+| Area | Files | Change |
+| --- | --- | --- |
+| subagent | projection.ts, projection-types.ts, index.ts | New `subagent` unit and its registration |
+| subagent | list-children.ts and its types | Rewritten as subagent-owned enumeration plus the projection-ladder four-state mapping; the session-query dependency, per-child event reads, and in-place classification machinery deleted; error code `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` replaced by `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`; new optional dependency dsh-session-projection-cache (pure read acceleration, skipped when absent) |
+| host/apiproxy | api-proxy.ts | `hasSubagentDescriptor` deleted; the owner check looks only at `header.origin`; `subagents.history` shares `session.history`'s source — live from in-memory events and the registry's watermark snapshot, cold from `inspectServable` reading persistence directly with a detached fold, no query service, the SESSION_QUERY_* error arms retired with it |
+| tool | tool-subagent-control/list-agents.ts | Load requirement narrowed (`sessionQuery` dropped from inject); model-visible schema, description, and rendering unchanged |
+| wire/client | api/subagents.ts, runtime sessions/service.ts, GUI | Types, row shape, and diagnostic handling **unchanged**; api/subagents.ts only reworded the `history` JSDoc to the dual arm |
+| core/session, session-persistence, session-projection(-cache), session-query(-sqlite) | — | **Zero changes** |
+
+## Alternatives considered
+
+**mode/label into SessionHeader.** The strongest zero-read guarantee — rows form from the header alone. But a header shape change propagates into both persistence backends and the header compatibility check; SQLite rejects pre-existing data outright, and JSONL pre-existing data can only degrade to unknown or be backfilled. Read-time computation's answer for pre-existing data is "one `inspect` computation on first listing", touching no durable format.
+
+**The projection-cache ladder (v3 draft: `cachedSnapshot ?? coldSnapshot` plus fail-soft write-back).** The mechanism works — session-projection-cache's checkpoint ladder is designed for cold reads in the first place. But checkpoint write-back is a whole list-driven body of derived-data persistence and invalidation orchestration (floor/identity/putSoft); what was rejected is that orchestration as the primary mechanism. The settled three-rung ladder later reuses this cache opportunistically, read-only, as its second rung — no write-back, no orchestration, skipped when absent.
+
+**A bounded-read primitive on persistence to rescue pre-existing data.** Opens a new seam primitive for a one-time problem; superseded by the read-time `inspect` full read — the full read the first time pre-existing data is listed is itself the value retrieval.
+
+**Optional mode/label on list rows (one v4 draft).** Healthy data is always computable; optionality merely spills garbage-data handling complexity onto every consumer — each consuming surface has to grow filter branches and an unknown display state. The strong contract plus omit-when-uncomputable is cleaner.
+
+**Deleting diagnostic rows outright (one v5 draft).** Deletion turns corpus-corruption visibility into rows silently vanishing, and wire/tool/GUI would each have to absorb contract and snapshot changes; retention only asks the list side to derive the classification from projection-value absence and activity, at zero cost. That damaged, dead child sessions in the corpus must be visible is the original motivation for diagnostics' existence, and with retention the consuming surfaces stay wholly unchanged.
+
+**A registry computation failure channel (per-unit fault tolerance plus a supplementary `failures` field).** To report corruption and unrecognized versions to consumers, we once considered having the registry catch unit exceptions and attach a per-key failure state beside the snapshot. Rejected: a failure is not a value and needs no channel — a unit never throws, absence is itself the signal, worst case the computation comes back empty, and how that is presented is the consumer's problem. The discussion of this route left one independent observation behind: the vendored Cordis `emit` ([vendor/cordis/src/events.ts](../../../../vendor/cordis/src/events.ts)) catches nothing a listener throws, so with the projection driver hanging off `session/event`, a unit exception would escape along emit — which adds weight to the "a unit never throws" discipline, but fixing emit fault tolerance is outside this note's scope.
+
+**Values landed with query index preparation (the v4/v5 settled design, built for a time).** Projection values folded into session index rows during the sqlite backend's reconciliation rebuild, for zero log reads in the steady read state; the `projectionsFor` bulk read face, the invalidation reconciliation of row values stored against the `(key → stateVersion)` registration set, and the SCHEMA bump were all actually built. Retired wholesale: the direction was backwards — query infrastructure was forced to learn domain vocabulary (projection columns, registration-set reconciliation) while the sole consumer, the subagent list, is satisfied by read-time computation; with consumers down to zero, this derived persistence has no reason to exist. `SESSION_QUERY_PROJECTIONS_UNAVAILABLE` was deleted along with the read face.
+
+**Subagent hand-rolled parsing plus an in-process memo plus creation seeding (v6 draft).** To excise the session-query dependency, we once considered the subagent package parsing descriptor events itself, avoiding repeated full reads with an in-process memo, and seeding initial values at creation. Superseded by the v7 ladder: live goes through the `sessionProjections` watermark cache and cold through `registry.restore`, reusing the registry's single fold authority — no second copy of descriptor-interpretation logic appears, and no process-state cache or seeding ordering is introduced.
+
+**DeepReadonly on the session-query output surface (a read-path overhaul experiment).** Make the public query outputs deeply readonly to pin immutable borrowing at the type level. Rejected on evidence: 3 TS2589 occurrences (excessively deep type instantiation) plus 17 sites of array-position contagion (consumers' array methods and spread sites forced to follow); deep immutability is guaranteed by core/session's runtime deep freeze, and that read-path overhaul is not part of this note.
+
+## Verification
+
+`packages/subagent/subagent/tests/list-children.spec.ts` is rewritten to this contract: live-only listing without persistence, query services, or the continuation runtime; with the registry absent, even zero children loudly report `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`; a live child incurs zero `inspect` throughout while a cold child incurs exactly one per listing; multiple descriptors resolve last-wins to the final one; corrupt payloads and unknown versions fold to `corrupt`; a cold-read failure maps to `unavailable` and retries on the next listing; the ancestor descriptor in a fork seed forms a row under that identity (pinning deviation one); ordinary forks and descendants without a subagent origin neither enter the list nor count toward `hasChildren`; `createdAt`-then-id ordering; an unmounted provider does not affect listing; compacted and uncompacted twins list identically; the three cases of pre-abort, persistence listing, and cold-read cancellation all normalize to `CANCELLED`; the empty list and stable error codes. A hostile-unit dual-path probe (`apply` lazily poisons, `view` detonates) proves that any registered unit's fold/schema throw on this child's log is contained as that child's `corrupt` row on both the live and the cold retrieval paths, with siblings and the listing itself unaffected. Second-rung cases: an own-seq identity used directly with zero `inspect`, a fork seed's ancestor identity (seq inside the seed range) rejected by the gate and falling through, an in-row identity absence (null sentinel or absent key) falling through, an absent cache service falling through, and a poisoned cache row silently falling through to the refold; cold-path lifecycle tampering degrades to `corrupt` field by witness field (`it.each` over the seven). The `tool-subagent-control` list-agents tests are updated for the narrowed load requirement; `optional-session-query.spec.ts` is deleted with the dependency it guarded; the existing keyless snapshots (`subagent-list-agents` among others) are unchanged, pinning that the healthy path's wire and model-visible surfaces did not move; a new keyless snapshot, `subagent-diagnostic` (examples/headless-agent), pins the four-state mapping's diagnostic classification — the model-visible changes such as descriptor-less settled debris becoming a `corrupt` row.
+
+## Consequences
+
+- Listing a live child reads zero log throughout; with the cache unmounted or missed, a cold child pays one full `inspect` read per listing, at a cost proportional to its transcript size and repeated with listing frequency — compute-and-discard is the settled stance: no cache of its own is built, nothing is written back, and short-term repeated full reads of the same id can hit the preparation-phase LRU, though listing does not depend on it.
+- The subagent list no longer requires a query backend: both pure-live and persistence-less deployments can list; `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` is gone, and loading the `list_agents` plugin no longer requires `sessionQuery`.
+- Identity interpretation exists only in the single unit registered with the registry: the list's three-rung ladder and GUI history's cold read all use the registry's and the cache's existing reads (snapshot, cachedSnapshot, restore), and no bypass fold exists; if some future consuming surface bypasses the registry with a hand-written fold, values will drift across read faces — a discipline this design requires be maintained, not a mechanical guarantee.
+- Per-child isolation is back: a single child's cold-read failure loses only that row and healthy siblings are unaffected; a persistence listing failure still fails the whole enumeration.
+- The diagnostic and enumeration semantics leaves six boundary deviations (a stillborn fork surfacing under its ancestor's identity, multiple descriptors resolving to the last, header conflicts going unnoticed, damaged-source read failures shifting from `corrupt` to `unavailable`, an unknown parent yielding an empty list instead of not-found, and rung 2's later-event window); the full semantics is in the known-boundary-deviations list; the first four are display or classification deviations on debris-grade data, the unknown-parent one is a silent query-semantics change, and the rung-2 window is a self-healing cache-serving divergence under the double condition of corruption plus a crash; resume authorization is unaffected throughout, all explicitly accepted.
+- Pre-#1569 data without `origin` is no longer recognized as a subagent owner; it never entered the catalog anyway, and pre-release carries no compatibility promise.
+
+## Related
+
+- [Durable subagent catalog and list_agents](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md) — partially superseded by this note: the descriptor remains the durable authority for mode/label and the fold input, while the list's enumeration and value retrieval move to the subagent-owned merge plus the projection ladder.
+- [Session projections and command lifecycle logging](../../proposed/architecture/2026-07-27-session-projection-and-command-log.md) — the authority for the registry contract; this note adds the `subagent` identity unit to it and becomes a consumer instance of the two existing reads, snapshot and restore.
+- [Web subagent conversations](../feature/2026-07-27-web-subagent-conversations.md) — the origin of `SessionHeader.origin` (#1569), the first half of taking identity determination off the log; its history cold read (inspect prefix plus registry fold) is the same-shape precedent for this note's value ladder.
+- [Reusable Session preparation before publication](2026-08-05-session-preparation.md) — the `inspect()` cold read and LRU reuse; the cold child's full-read cost model builds on it.

.agents/notes/implemented/architecture/2026-08-06-subagent-list-identity-projection.zh.md

@@ -0,0 +1,184 @@
+# Agent Note: subagent 列表经投影单元读取身份
+
+Status: implemented
+
+[English](2026-08-06-subagent-list-identity-projection.md) | 中文
+
+## 问题
+
+重写前的 `SubagentService.listChildren` 对每个 `header.origin === 'subagent'` 的直接 child，每次列表都执行 `listEvents` 加 `readEvent` 两次整日志物化，且每次物化都伴随整日志 structuredClone，只为从描述符事件里折出 mode 与 label 两个字段。描述符在日志中的位置不固定——fork 前缀任意长，zstd 压缩帧没有 seq 索引——因此定位没有捷径；这条路径没有任何缓存，代价随 transcript 长度 × child 数量 × 列表频率放大。它还把 session-query 拉成列表的硬依赖：没有 query backend 的部署，`list_agents` 以 `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` 整体拒绝，尽管枚举所需只是 header 事实。
+
+同一根因还有第二个症状：host 侧的 `hasSubagentDescriptor()` 在每次 Agent 绑定 RPC 的属主判定上扫描目标会话的 own suffix，即便 `SessionHeader.origin` 已经回答了同一个问题的绝大部分。
+
+根因在于 [durable-subagent-catalog 决策](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md)把描述符事件（`subagent/descriptor`）定为目录的唯一持久权威，却没有为描述符读取配任何缓存层，并把逐 child 双读明确接受为"无索引的正确性基线"。[web subagent conversations](../feature/2026-07-27-web-subagent-conversations.md)（#1569）已把"是不是 subagent"放进了 header（`SessionHeader.origin`），身份判定不再读日志；mode 与 label 仍然要扫。
+
+## 决策
+
+mode 与 label 由新的 `subagent` projection unit（纯身份两臂）折叠，unit 是折叠规则的唯一权威；`listChildren` 不再依赖 session-query——枚举是 subagent 自管的 live-preferred 合并，取值走三级"算完即止"阶梯：live child 同步读注册表的既有水位缓存（零日志读）；cold child 先问可选的 `sessionProjectionCache` checkpoint，取到过 seq 门的身份即定值；否则一次 `persistence.inspect` 整读加 `registry.restore` 折叠。无索引、不自建缓存、无回写。
+
+消除逐 child 扫描的出路有三类：把 mode/label 提升进 header（写路承担）；为投影建持久派生（checkpoint 阶梯，或随查询索引重建落值、读端对账）；读时现算（live 走水位缓存，cold 一次整读）。本记录取第三条。"值随查询索引落库"曾是本记录的定稿方向并一度施工，最终整体退役：查询基础设施被迫认识领域词汇，而唯一消费方读时现算即可满足——live child 的零读由 session-projection 既有水位缓存白拿，cold child 的一次整读被"算完即止"显式接受。前两条与退役理由详见考虑过的替代方案一节。
+
+要点：
+
+- **subagent 列表不依赖 session-query**：枚举由 subagent 自管的 live-preferred 合并完成，mode/label 经 `ctx.sessionProjections` 取值；没有 query backend 的部署照常列表。
+- **取值三级"算完即止"阶梯**：live child 读 `sessionProjections.snapshot()`（注册表既有水位缓存，零日志读）；cold child 先读可选 `sessionProjectionCache.cachedSnapshot(header)`，values 含非 null 且过 seq 门（`seq >= seedLength ?? 0`）的 `subagent` 身份即直接用；否则一次 `persistence.inspect` 整读加 `registry.restore({}, events, 0)` 折叠；再没有就没有——不自建缓存、无回写、无索引。
+- **`subagent` projection unit 是折叠规则唯一权威**：live snapshot、cold restore、GUI history 的 detached 折叠全部经 registry 计算，不存在第二份描述符解释逻辑。
+- **header、描述符（v2）、session-persistence、session-projection(-cache)、session-query(-sqlite) 全部零改动**；存量数据第一次被列表时一次 `inspect` 现算获得精确值，无 unknown 降级态、无迁移。
+
+与既有记录的关系：
+
+- 本记录取代 [durable-subagent-catalog](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md) 中列表读路径的两项设计：经 `sessionQuery.traceSession` 枚举，与逐 child 读取描述符事件（`listEvents` 加精确 `readEvent` 双读、就地诊断分类）。diagnostic 行语义保留，分类改由列表按投影值缺席与 activity 派生；描述符事件仍是 mode/label 的唯一持久权威与折叠输入，恢复鉴权与激活契约不动。属部分取代，两记录保持交叉链接。
+- [session-projection RFC](../../proposed/architecture/2026-07-27-session-projection-and-command-log.md) 的 registry 契约（`ProjectionDefinition`、`snapshot`、`restore`）零改动，本记录只为其新增 `subagent` 身份 unit 一个注册项，并成为 snapshot（live）与 restore（cold）两处既有读法的又一消费实例——GUI history 的冷读已是同款。折叠规则只在 registry 注册一份；任何消费面都经 registry 计算，不存在第二份折叠逻辑。
+
+### `subagent` projection unit
+
+挂在现有 `subagentTiming` 旁（[projection.ts](../../../../packages/subagent/subagent/src/projection.ts)、[projection-types.ts](../../../../packages/subagent/subagent/src/projection-types.ts)），key 为 `subagent`：
+
+```ts ignore-check
+export type SubagentIdentityProjection =
+  | { mode: 'one-shot'; label?: string; seq: number }
+  | { mode: 'continuable'; label: string; seq: number }
+
+declare module '@deepseek-ai/dsh-session-projection/types' {
+  interface SessionProjectionMap {
+    subagent: SubagentIdentityProjection | null
+  }
+}
+```
+
+- 投影是纯身份，**projection 体系不做失败通道**：unit 永不抛错；载荷损坏、版本不认识与整日志没有描述符一样，折叠结果是**可序列化的 null 哨兵**——map 条目为 `SubagentIdentityProjection | null`，非可选、非 undefined/缺 key。理由：registry 的 onChanged 推送经 JSON 序列化，undefined 字段被 stringify 丢弃，客户端帧校验拒收，消费方存储的旧身份将永不更新；null 完好过帧，消费方以哨兵替换旧身份。判定纪律：消费面把 null 与 undefined（仅 JSON 边界丢 key 可产生）一律视为无值。"算出来没有"如何呈现是消费方自己的事（见下文 `listChildren` 四态映射）。
+- label 强度由描述符 schema 决定：continuable 的 label 解析强制必有，one-shot 的本就可选；mode/label 判别与下文 child 行的强契约完全一致（行不携带 `seq`——它是投影内部的 own-suffix 证明）。
+- 身份携带 `seq`：折出该身份的 `subagent/descriptor` 事件 seq，两臂必有、null 哨兵无——`seq >= header.seedLength ?? 0` 证明身份折叠自 child 自身后缀，而非 fork 种子回放的祖先描述符。state 增 `seq` 使 unit `stateVersion` 升至 2，既存 checkpoint 行按 registry 契约版本失配失效、落权威重折。
+- 折叠规则：`subagent/descriptor` last-wins，与 `subagentTiming` 同一条 descriptor-reset 纪律——fork 前缀里的祖先描述符被自身描述符覆盖。损坏或版本不认识的载荷同样 last-wins：重置为 null 哨兵而非保留先前身份，健康祖先的 fork 不会继承自身描述符立不住的身份。
+
+### 枚举：subagent 自管 live-preferred 合并
+
+`listChildren`（[list-children.ts](../../../../packages/subagent/subagent/src/list-children.ts)）的枚举不经任何查询服务：`ctx.sessions.list()` 与 `ctx.get('sessionPersistence')?.list()` 两个来源按 id 合并，live 记录整条覆盖同 id 持久化记录、不做 header 一致性校验。枚举所需全部是 header 事实：
+
+- 过滤：`header.origin === 'subagent' && header.parentSession === parentSessionId`。
+- `hasChildren`：同一份合并材料向下看一层——存在 `origin === 'subagent'` 且 `parentSession` 为该 child 的直接后代。
+- `activity`：live 记录为 `running`，仅存在于持久化的为 `inactive`。
+- 排序：`createdAt` 升序、再按 child id 升序（与旧契约一致）。
+- **persistence 缺席退为 live-only 枚举，不报错**：没有 persistence 的部署，cold child 本就无法 resume，列出 live child 仍然有意义。（对照：旧实现在 sessionQuery 缺失时整体拒绝。）
+- persistence 列表失败使整次枚举失败；per-child 隔离只作用于逐 child 的冷读。
+
+### 取值：三级"算完即止"阶梯
+
+对每个枚举出的 child，mode/label 取值走三级阶梯——算完即止，不自建缓存、无回写（第三级与 apiproxy `session.history` 的冷读同款）：
+
+| 级 | 读法 | 成本 |
+| --- | --- | --- |
+| 1：live child | `ctx.sessionProjections.snapshot(session).values.subagent` | 零日志读——注册表既有水位缓存，同步取值 |
+| 2：cold child，cache 命中 | 可选 `sessionProjectionCache.cachedSnapshot(header)`，values 含非 null 的 `subagent` 身份且 `identity.seq >= header.seedLength ?? 0` 才直接用——own descriptor 一经追加不可变，seq 门证明该值折叠自 child 自身后缀，无视行水位 | 零日志读 |
+| 3：cold child，兜底 | `persistence.inspect(id)` 整读 + `registry.restore({}, events, 0).snapshot.values.subagent` | 每次列表一次整读现算 |
+
+- 错误契约：`ctx.sessionProjections` 未挂载是配置错误，`listChildren` 在枚举前无条件检查并以 `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE` 响亮失败——零 children 的部署同样确定失败，不因列表恰好为空而掩盖配置问题。会话存储同理：`ctx.get('sessions')`（严格全局读取，不走调用方作用域的属性代理）缺席以 `SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE` 失败。两码的 wire 映射有别：apiproxy 只为 `PROJECTIONS_UNAVAILABLE` 设专门 wire 脸，`SESSION_STORE_UNAVAILABLE` 走通用 internal 兜底——apiproxy 组合自身就 inject `sessions`，该错误在其部署不可达，专门映射违反 need 原则。`SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` 已随 session-query 依赖删除。
+- cache 是纯可选加速层：服务缺席判空跳过——无错误码、不进配置校验（与 `sessionProjections` 的响亮契约相对）。第二级任何抛错（包括缓存内任一 unit 行中毒使 `viewCheckpoint` 引爆）静默落第三级——缓存是派生数据，其故障不产生 `corrupt` 判决，终审归权威重折；checkpoint 切面早于描述符的行，`subagent` key 天然缺席，自动落底，无特判；行里的 null 哨兵同样不作数——一律落第三级，由权威重折裁决。创建窗口内的 count/interval checkpoint 可能把 fork 种子回放的祖先身份落进行——祖先 seq 落在 seed 区间，被 seq 门拒绝，同样落第三级裁决。
+- per-child 隔离：单 child 的 cold 整读失败只使该行成为 `unavailable` diagnostic，下次列表自然重试，不影响 sibling（见四态映射）。
+- 冷路径的生命周期见证：preparation 的结果必须仍指向枚举时的那个生命周期——见证字段集与旧 SOURCE_CONFLICT 检查同款七字段（version、id、createdAt、cwd、parentSession、seedLength、delegationDepth）；同 id 删除后重新发布的会话对旧 parent 的目录降级为 `corrupt` 行，不外漏新 owner 的 child。
+- 冷读并发以常数 4 有界——它约束的是本地介质的一次只读扫描而非部署行为；出现联网 persistence backend 时提升为验证过的 `Config` 字段。
+- 冷读成本如实记录：cache 未挂载或未命中时，cold child 每次列表才付一次整读，成本与其 transcript 大小成正比；定案"算完即止"，不自建缓存。整读经 `inspect()` 走 [Session 准备阶段](2026-08-05-session-preparation.md)的冷读，同 id 短期重复读取可命中其 LRU 复用，但列表不依赖此。live child 全程零日志读。
+- 取消：每次 persistence 读前后检查调用方 signal，abort 之后才结算的读拒绝归一化为稳定错误码 `CANCELLED`。
+
+### 权威模型
+
+- session log 是唯一权威；本方案不新增任何派生持久化——没有索引值、没有自己的 checkpoint、没有进程 memo；第二级读取的 `sessionProjectionCache` checkpoint 是既有组合项的派生数据，本方案只读不写。取值现算现弃，值的新鲜度就是读取时点的 live 状态或持久化 revision（own descriptor 一经追加不可变——缓存身份过 seq 门后无陈旧性问题，门防的是种子回放的祖先身份）。
+- Session 与 persistence 写路完全不感知列表与投影消费：没有事件监听回写，没有写时折叠。
+- 枚举与取值不构成第二个鉴权来源，也不让尚未发布的 child 可见——两个来源只见已发布的 live 记录与已落盘的持久化记录，与 durable-subagent-catalog 记录对派生读面立下的规则一致。
+
+### `listChildren` 行形状与消费面
+
+`SubagentListEntry` **数据结构与重写前完全一致**——child 与 diagnostic 两臂、`kind` 判别、reason 三值、child 臂的 mode/label 强契约全部保留；变化只在诊断的信息来源：投影体系没有失败通道，diagnostic 由列表按投影值缺席与 activity 派生，列表本身零事件解析。"没有就等待硬读取"保证阶梯对健康数据必然算得出 mode/label。
+
+```ts ignore-check
+export type SubagentListEntry =
+  | ({
+    readonly kind: 'child'
+    readonly id: SessionId
+    readonly activity: 'running' | 'inactive'
+    readonly hasChildren: boolean
+  } & (
+    | { readonly mode: 'one-shot'; readonly label?: string }
+    | { readonly mode: 'continuable'; readonly label: string }
+  ))
+  | {
+    readonly kind: 'diagnostic'
+    readonly id: SessionId
+    readonly reason: 'corrupt' | 'unsupported' | 'unavailable'
+  }
+```
+
+对每个枚举出的 child，阶梯取值结果按四态映射成行：
+
+| 阶梯取值结果 | 行 |
+| --- | --- |
+| 快照含非 null 的 `subagent` 身份 | child 行 |
+| 快照在、`subagent` 为 null 哨兵或 key 缺席，且 child **inactive** | diagnostic 行，reason `corrupt`（定局残骸：无、损坏或版本不认识的描述符，不再细分） |
+| 快照在、`subagent` 为 null 哨兵或 key 缺席，且 child **running** | 行不出现（创建窗口：描述符尚未追加，与旧实现同窗口 omit） |
+| cold 整读失败 | diagnostic 行，reason `unavailable` |
+
+- `unsupported` 不再被产出：类型与 wire 枚举按"数据结构保持现状"留存该成员，本记录留档其为不再产出。
+- descriptor-less 定局残骸从旧实现的 omit 归入 `corrupt` diagnostic——库里的坏、死子会话可见，不静默消失，这正是保留 diagnostic 的原始动机。
+- 任一注册 unit 的 fold/schema 在该 child 日志上抛错，同样收纳为该 child 的 diagnostic 行，reason `corrupt`——确定性数据故障，对齐旧实现 `SESSION_QUERY_CORRUPT_SESSION`→`corrupt` 的映射语义；live 与 cold 同待遇，逐 child 隔离，sibling 与列表本身不受影响。它与「无值 + running → omit」正交：创建窗口是"尚无数据"，fold 抛错是"数据坏了"——running 的中毒 child 也出 `corrupt` 行而非 omit。
+
+已知边界偏差（有意接受，随本记录留档）：
+
+- 死于发布窗口的 fork child，seed 里若有祖先描述符，last-wins 会给出祖先身份，误现为 child 行；恢复仍按 own-suffix 折叠权威失败（`NOT_RESUMABLE`）。旧实现靠 `seedLength` 过滤将其 omit；projection unit 看不到 header，接受此残骸级偏差（`subagentTiming` 有同类既有暴露）。
+- own suffix 出现多个描述符，旧实现判 corrupt，现 last-wins 取末者（provider 契约本就保证恰一）。
+- live/persisted header 冲突，旧实现是 per-child corrupt；现枚举 live 优先、不做一致性校验，冲突不再被察觉，以 live 记录成行。
+- 损坏存储的源读失败（如坏 surface 被冷读整读拒收），旧实现映射 per-child `corrupt`，现统一成 `unavailable` 行（读侧无从区分成因）。
+- 未知 parent，旧实现经 session-query 抛 not-found（'parent session … was not found'）；现自管合并对不存在的 parent 得到空子集，枚举返回空列表，wire 上后续操作落到 child 级 subagent-not-found——语义与文案的静默变化，显式接受。
+- rung 2 的更晚事件窗口：cache 行恰在首个自有描述符之后落盘，日志随后追加第二个自有描述符（或 malformed 载荷置 null 哨兵），且进程在下一次 checkpoint 前崩溃——此后冷列表的 rung 2 凭 seq≥seedLength 门持续供出行内旧身份（第一个自有描述符的值），与权威重折（last-wins 第二个）分歧，且 rung 2 命中期间不触发重折、无从察觉。边界三条：①前提是同一 child 出现第二个自有描述符，违反 establishing provider"恰追加一次"契约，属损坏类数据，与多描述符偏差同族同源；②需"损坏 + 崩溃错过 checkpoint（turn/end 与 disposal 两个 mandatory 点及 count/interval 节流点全部未及）"双条件同时成立；③健康 child（恰一自有描述符）不受影响——seq 门放行的正是唯一真身份。自愈条件：该 child 任一次 live 运行（turn/end mandatory checkpoint）或任何触发 cache.write 的时点，都会以新 fold 整行覆写（whole-record replace），rung 2 随即供正；权威路径（rung 3 重折、live snapshot、resume 折叠）自始正确，分歧只存在于持续冷、行未再更新期间的列表读。机制修法不采：gate 对账需知日志末端 seq，冷路径零读不可得；cache 行携 revision 是 opaque token，无法比较且跨域改 schema——按"cache 永不为权威"总纲归档为接受项。
+
+消费面：wire、tool、GUI 的 diagnostic 处理**全部保持原状零改动**（`list_agents` 的 description 与 output schema 未动；该插件仅加载要求收窄——inject 去掉 `sessionQuery`）。行为上动的只有 apiproxy：路由段的 `hasSubagentDescriptor()` 扫描已删除，`hasSubagentOwner` 只看 `header.origin`——pre-#1569 的无 `origin` 存量不再被认作 subagent 属主，其本就不进目录，pre-release 立场接受；`subagents.history` 与 `session.history` 同源对齐——live child 用内存事件与注册表水位快照，cold child 用 `inspectServable` 直读持久化并 detached 折叠，不经查询服务，SESSION_QUERY_* 错误臂随之退役，wire 形状不变（`history` 的 JSDoc 措辞改为 live 内存快照／cold 持久日志双臂）。
+
+### 改动落点
+
+| 区域 | 文件 | 改动 |
+| --- | --- | --- |
+| subagent | projection.ts、projection-types.ts、index.ts | 新 `subagent` unit 与注册 |
+| subagent | list-children.ts 及类型 | 重写为自管枚举 + 投影阶梯四态映射；删 session-query 依赖、逐 child 事件读取与就地分类机器；错误码 `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` 换 `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`；新增可选依赖 dsh-session-projection-cache（纯加速读取，缺席跳过） |
+| host/apiproxy | api-proxy.ts | 删 `hasSubagentDescriptor`，属主判定只看 `header.origin`；`subagents.history` 与 `session.history` 同源——live 用内存事件与注册表水位快照，cold 用 `inspectServable` 直读持久化并 detached 折叠，不经查询服务，SESSION_QUERY_* 错误臂随之退役 |
+| tool | tool-subagent-control/list-agents.ts | 加载要求收窄（inject 去 `sessionQuery`）；model-visible schema、描述与渲染零改动 |
+| wire/client | api/subagents.ts、runtime sessions/service.ts、GUI | 类型、行形状与 diagnostic 处理**零改动**；api/subagents.ts 仅 `history` 的 JSDoc 措辞改为双臂 |
+| core/session、session-persistence、session-projection(-cache)、session-query(-sqlite) | — | **零改动** |
+
+## 考虑过的替代方案
+
+**mode/label 进 SessionHeader。** 零读保证最强——列表只看 header 就能成行。但 header 形状变更传导两个 persistence backend 与 header 兼容检查；SQLite 存量直接拒收，JSONL 存量只能 unknown 降级或 backfill。读时现算对存量的答案是"第一次列表一次 `inspect` 现算"，不碰持久格式。
+
+**projection-cache 阶梯（v3 稿：`cachedSnapshot ?? coldSnapshot` 加 fail-soft 写回）。** 机制成立——session-projection-cache 的 checkpoint 阶梯本就为冷读设计。但 checkpoint 写回是一套由列表驱动的派生数据持久化与失效编排（floor/identity/putSoft）；被否的是这套编排作为主机制。定稿的第三级阶梯后来以只读方式机会性复用该缓存作第二级——无写回、无编排、缺席即跳过。
+
+**给 persistence 加有界读原语抢救存量。** 为一次性问题新开 seam 原语；被读时 `inspect` 整读取代——存量第一次被列表时的整读就是取值本身。
+
+**list 行 mode/label 可选化（v4 一稿）。** 健康数据必然可算；可选化只是把垃圾数据的处理复杂度外溢给全部消费方——每个消费面都要长出过滤分支和 unknown 展示态。强契约加算不出即 omit 更干净。
+
+**彻底删除 diagnostic 行（v5 一稿）。** 删除把库损坏的可见性外溢为行静默消失，wire/tool/GUI 反要各自承担契约与快照变更；而保留只需列表侧按投影值缺席与 activity 派生分类，零成本。库里的坏、死子会话必须可见是 diagnostic 存在的原始动机，保留后消费面整体零改动。
+
+**registry 计算失败通道（per-unit 容错加 `failures` 附加字段）。** 为把损坏、版本不认识报告给消费方，曾考虑让 registry 捕获 unit 异常并在 snapshot 旁附 per-key 失败态。被否：failure 不是值，也不必是通道——unit 永不抛错，缺席本身就是信号，"大不了算出来没有"，如何呈现是消费方要考虑的事。该路线讨论顺带留下一个独立观察：vendor cordis 的 `emit`（[vendor/cordis/src/events.ts](../../../../vendor/cordis/src/events.ts)）对 listener 抛错零捕获，投影驱动挂在 `session/event` 上时 unit 异常会沿 emit 逃逸——这加重了"unit 永不抛错"纪律的分量，但 emit 容错的修复不属于本记录范围。
+
+**值随 query 索引 preparation 落库（v4/v5 定稿，一度施工）。** 投影值在 sqlite backend 的对账重建里折叠落进 session 索引行，读稳态零日志；`projectionsFor` 批量读面、行值随 `(key → stateVersion)` 注册集存储的失效对账与 SCHEMA bump 均已施工过。整体退役：方向反了——查询基础设施被迫认识领域词汇（投影列、注册集对账），而唯一消费方 subagent 列表读时现算即可满足；消费方归零后，这套派生持久化没有存在理由。`SESSION_QUERY_PROJECTIONS_UNAVAILABLE` 随读面一并删除。
+
+**subagent 手工 parse 加进程 memo 加创建播种（v6 稿）。** 为摘除 session-query 依赖，曾考虑 subagent 自己解析描述符事件、以进程内 memo 避免重复整读、创建时播种初值。被 v7 阶梯取代：live 走 `sessionProjections` 水位缓存、cold 走 `registry.restore`，复用 registry 这一份折叠权威，不再出现第二份描述符解释逻辑，也不引入进程态缓存与播种时序。
+
+**session-query 输出面 DeepReadonly（读路径改造实验）。** 公开查询输出深只读化，以在类型层面钉死不可变借用。实证否决：3 处 TS2589（类型实例化过深）加 17 处数组位传染（消费方数组方法与展开处被迫跟改）；深层不可变由 core/session 的运行时深冻结保证，该读路径改造未纳入本记录。
+
+## 验证
+
+`packages/subagent/subagent/tests/list-children.spec.ts` 重写为本契约：无 persistence、query 服务与继续运行时的 live-only 列表；registry 缺席时零 children 也响亮报 `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`；live child 全程零 `inspect`、cold child 每次列表恰一次；多描述符 last-wins 取末者；损坏载荷与未知版本折为 `corrupt`；冷读失败映射 `unavailable` 且下次列表重试；fork seed 里的祖先描述符按该身份成行（偏差一钉住）；普通 fork 与无 subagent origin 的后代不入列也不计入 `hasChildren`；`createdAt`→id 排序；provider 未挂载不影响列表；压缩与未压缩孪生一致；预中止、持久化列表与冷读取消三例归一 `CANCELLED`；空列表与稳定错误码。敌意 unit 双路探针（`apply` 惰性置毒、`view` 引爆）证明任一注册 unit 在该 child 日志上的 fold/schema 抛错，在 live 与 cold 两条取值路径上都收纳为该 child 的 `corrupt` 行，sibling 与列表本身不受影响。第二级例：own-seq 身份直用零 `inspect`、fork 种子祖先身份（seq 落在 seed 区间）被门拒绝落底、行内无身份（null 哨兵或 key 缺席）落底、cache 服务缺席落底、缓存行中毒静默落底重折；冷路径 lifecycle 篡改按见证七字段逐一（`it.each`）降级为 `corrupt`。`tool-subagent-control` 的 list-agents 测试随加载要求收窄更新；`optional-session-query.spec.ts` 随依赖消失删除；既有无密钥快照（`subagent-list-agents` 等）零变化，钉住健康路径的 wire 与 model-visible 面不变；新增无密钥快照 `subagent-diagnostic`（examples/headless-agent）钉住四态映射的诊断分类——descriptor-less 定局残骸成 `corrupt` 行等模型可见变化。
+
+## 后果
+
+- live child 的列表全程零日志读；cold child 在 cache 未挂载或未命中时每次列表一次 `inspect` 整读，成本与其 transcript 大小成正比、随列表频率重复——定案"算完即止"，不自建缓存、不回写，同 id 短期重复整读可命中准备阶段 LRU 但列表不依赖它。
+- subagent 列表不再要求 query backend：纯 live 与无 persistence 的部署都能列表；`SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` 消失，`list_agents` 插件加载不再要求 `sessionQuery`。
+- 身份解释只存在于 registry 注册的一份 unit：列表三级阶梯与 GUI history 冷读走的都是 registry 与 cache 的既有读法（snapshot、cachedSnapshot、restore），不存在旁路折叠；若未来某消费面绕开 registry 手写折叠，各读面的值将漂移——这是本设计要求维持的纪律，不是机制保证。
+- per-child 隔离回归：单 child 冷读失败只损失该行，healthy sibling 不受影响；persistence 列表失败仍使整次枚举失败。
+- 诊断与枚举语义留下六处边界偏差（stillborn fork 祖先身份误现、多描述符取末者、header 冲突不再被察觉、损坏源读失败由 `corrupt` 转 `unavailable`、未知 parent 由 not-found 改为空列表、rung 2 更晚事件窗口），完整语义见已知边界偏差清单；前四处为残骸级数据的展示或分类偏差，未知 parent 一处是查询语义的静默变化，rung 2 窗口一处是损坏加崩溃双条件下可自愈的缓存供值分歧；恢复鉴权均不受影响，显式接受。
+- pre-#1569 的无 `origin` 存量不再被认作 subagent 属主；其本就不进目录，pre-release 无兼容承诺。
+
+## 相关
+
+- [durable-subagent-catalog 与 list_agents](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md)——被本记录部分取代：描述符仍是 mode/label 的持久权威与折叠输入，列表的枚举与取值改为自管合并加投影阶梯。
+- [session projections 与命令生命周期日志](../../proposed/architecture/2026-07-27-session-projection-and-command-log.md)——registry 契约的权威；本记录为其新增 `subagent` 身份 unit，并成为 snapshot/restore 两处既有读法的消费实例。
+- [web subagent conversations](../feature/2026-07-27-web-subagent-conversations.md)——`SessionHeader.origin` 的出处（#1569），身份判定去日志化的前半步；其 history 冷读（inspect 前缀加 registry 折叠）是本记录取值阶梯的同款先例。
+- [发布前可复用的 Session 准备阶段](2026-08-05-session-preparation.md)——`inspect()` 冷读与 LRU 复用；cold child 整读的成本模型建立其上。

.agents/notes/implemented/bug-fix/2026-07-28-web-agent-runtime-context.i18n.yaml

@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-07-28-web-agent-runtime-context.md
-2026-07-28-web-agent-runtime-context.md: 449c9d4ba2b144d02dee4b98ae80c86815aec5c1
-2026-07-28-web-agent-runtime-context.zh.md: def1674be5f193739bfb214a24f34590ee075d5f
+2026-07-28-web-agent-runtime-context.md: 6c2bb8faf0db56d2a29b47726ab295c3501cce2d
+2026-07-28-web-agent-runtime-context.zh.md: 09cbbdeabc244c265381b1ad67c3629b4c0763f5

.agents/notes/implemented/bug-fix/2026-07-28-web-agent-runtime-context.md

@@ -10,13 +10,13 @@ The shared CLI base configured an empty deployment persona, the Web overlay did
 
 ## Decision
 
-The shared Web/headless overlay (`apps/cli/config/web.cordis.yml`) supplies a concise coding-agent persona containing the resolved `{{model}}` and session `{{cwd}}`. `dsh web` additionally resolves the harness checkout from the launcher's module URL, installs the existing `harness:source` section, and adds an `app:web-surface` section before serving requests. The launcher registers that setup before mounting the config tree; its `systemPrompt` injection therefore installs both sections before later prompt consumers such as the agent loop can activate and emit a request header. The [source-checkout/workdir decision](2026-07-30-source-checkout-workdir-distinction.md) owns the source section's wording and its warning not to infer one path from the other.
+The Web profile composes the `dsh-base` and `dsh-web-app` bundles. The Web bundle supplies a concise coding-agent persona containing the resolved `{{model}}` and session `{{cwd}}`; its `web-runtime` plugin adds the `app:web-surface` section when `surfaceContext` is true. Before mounting the profile tree, the `dsh web` alias reads that same composed setting and installs the existing `harness:source` section only when the surface context is enabled. The headless bundle and a profile that owns its complete prompt set `surfaceContext: false`, suppressing the Web prompt and managed shell facts; the Web alias also suppresses the source section without checking an overlay path. Every mounted prompt contribution still activates before consumers such as the agent loop can emit a request header. The [source-checkout/workdir decision](2026-07-30-source-checkout-workdir-distinction.md) owns the source section's wording and its warning not to infer one path from the other.
 
 The Web section treats unqualified references to “this page,” “this GUI,” or “this app” as references to the DeepSeek Harness Web GUI. It also states that the browser provides no implicit DOM, route, or screenshot context, so the model can identify the product without claiming visual state it did not receive. The assembled text is logged in `request/header`, preserving the model-visible/logged invariant.
 
 ## Verification
 
-The focused startup-order test registers a later `systemPrompt` consumer and proves that it observes both launcher sections on its first activation. The keyless fresh-round-trip Web scenario boots the shipped base plus Web overlay, registers the same launcher context as `dsh web`, runs a real session through the HTTP/SSE application, and snapshots the first four system-prompt sections with source and working-directory paths normalized. The snapshot pins the harness identity, source checkout, Web orientation, and resolved coding-agent persona in request order.
+The Web runtime unit tests pin both enabled and disabled `surfaceContext` behavior, while the Web alias unit test pins default-on and explicit-off source-section gating from the composed row. The keyless fresh-round-trip Web scenario boots the shipped base plus Web bundle, runs a real session through the HTTP/SSE application, and snapshots the system-prompt prefix with source and working-directory paths normalized. The snapshot pins the harness identity, source checkout, Web orientation, and resolved coding-agent persona in request order. The Core Web snapshot applies the RL overlay and pins its complete system prompt with no source or Web section.
 
 ## Alternatives considered
 
@@ -30,4 +30,4 @@ The focused startup-order test registers a later `systemPrompt` consumer and pro
 
 ## Consequences
 
-Web requests gain a short stable prompt prefix and may invalidate provider prefix caches once when this change is deployed. Agents can distinguish the GUI source checkout from the selected Workspace and resolve ordinary references to the current app without a clarification round trip. References to a specific visual state remain bounded by the explicit no-DOM/no-route/no-screenshot statement and may still require a path, description, or attachment.
+Ordinary Web requests gain a short stable prompt prefix and may invalidate provider prefix caches once when this change is deployed. Agents can distinguish the GUI source checkout from the selected Workspace and resolve ordinary references to the current app without a clarification round trip. References to a specific visual state remain bounded by the explicit no-DOM/no-route/no-screenshot statement and may still require a path, description, or attachment. Complete-prompt profiles can opt out through the Web runtime's composition setting without a launcher path check.

.agents/notes/implemented/bug-fix/2026-07-28-web-agent-runtime-context.zh.md

@@ -10,13 +10,13 @@ CLI 共享 base 配置了空的部署 persona，Web overlay 没有替换它，
 
 ## 决策
 
-`apps/cli/config/web.cordis.yml` 这份 Web／无头共享 overlay 提供一段简洁的编码 agent persona，其中包含解析后的 `{{model}}` 与会话 `{{cwd}}`。`dsh web` 还会根据启动器模块的 URL 解析 harness checkout，安装现有的 `harness:source` 提示词段，并在对外提供请求服务前添加 `app:web-surface` 提示词段。启动器会在挂载配置树前注册这项设置；因此，它的 `systemPrompt` 注入会在 agent loop（智能体循环）等后续提示词消费方激活并发出 request header 之前安装这两个提示词段。源码提示词段的措辞，以及其中不得从一条路径推断另一条路径的警告，均由另行记录的[源码 checkout 与工作目录区分决策](2026-07-30-source-checkout-workdir-distinction.md)负责。
+Web profile 会组合 `dsh-base` 与 `dsh-web-app` 两个组合包。Web 组合包提供一段简洁的编码 agent persona，其中包含解析后的 `{{model}}` 与会话 `{{cwd}}`；当 `surfaceContext` 为 true 时，其 `web-runtime` 插件会添加 `app:web-surface` 提示词段。挂载 profile 配置树前，`dsh web` 别名会读取组合后的同一项设置，并且仅在启用表层上下文时安装现有的 `harness:source` 提示词段。无头组合包和拥有完整提示词的 profile 都会设置 `surfaceContext: false`，从而抑制 Web 提示词与受管 shell 事实；Web 别名也会抑制源码提示词段，而无需检查 overlay 路径。每项已挂载的提示词贡献仍会在 agent loop（智能体循环）等消费方发出 request header 前激活。源码提示词段的措辞，以及其中不得从一条路径推断另一条路径的警告，均由另行记录的[源码 checkout 与工作目录区分决策](2026-07-30-source-checkout-workdir-distinction.md)负责。
 
 Web 提示词段把未限定的「这个页面」「这个 GUI」或「这个应用」解释为 DeepSeek Harness Web GUI。同时，它会明确说明浏览器不会隐式提供 DOM、路由或截图上下文，使模型能够识别产品，但不会声称掌握未收到的视觉状态。组装后的文本会记录在 `request/header` 中，从而保持「模型可见内容必须有日志记录」这一不变量。
 
 ## 验证
 
-聚焦启动顺序的测试会注册一个后续的 `systemPrompt` 消费方，并证明该消费方首次激活时就能观察到启动器的两个提示词段。无密钥的 Web fresh-round-trip 场景会启动已交付的 base 与 Web overlay，注册与 `dsh web` 相同的启动器上下文，并通过 HTTP／SSE 应用运行一个真实会话。测试会把源码路径和工作目录规范化，然后对系统提示词的前四个段落生成快照。该快照按请求顺序固定 harness 身份、源码 checkout、Web 界面定位，以及解析后的编码 agent persona。
+Web 运行时单元测试会固定启用和禁用 `surfaceContext` 时的行为，Web 别名单元测试则会固定组合后配置行对源码提示词段的默认启用与显式禁用门控。无密钥的 Web fresh-round-trip 场景会启动已交付的 base 与 Web 组合包，通过 HTTP／SSE 应用运行一个真实会话，并在规范化源码路径与工作目录后对系统提示词前缀生成快照。该快照按请求顺序固定 harness 身份、源码 checkout、Web 界面定位，以及解析后的编码 agent persona。Core Web 快照会应用 RL overlay，并固定不含源码提示词段或 Web 提示词段的完整系统提示词。
 
 ## 考虑过的替代方案
 
@@ -30,4 +30,4 @@ Web 提示词段把未限定的「这个页面」「这个 GUI」或「这个应
 
 ## 影响
 
-Web 请求会增加一段较短且稳定的提示词前缀；部署此变更时，模型提供方的前缀缓存可能失效一次。agent 可以区分 GUI 源码 checkout 与所选 Workspace，并且无需再经过一轮澄清即可解析对当前应用的一般指代。对特定视觉状态的指代仍受「无 DOM／无路由／无截图」这一显式边界约束，必要时仍需用户提供路径、描述或附件。
+常规 Web 请求会增加一段较短且稳定的提示词前缀；部署此变更时，模型提供方的前缀缓存可能失效一次。agent 可以区分 GUI 源码 checkout 与所选 Workspace，并且无需再经过一轮澄清即可解析对当前应用的一般指代。对特定视觉状态的指代仍受「无 DOM／无路由／无截图」这一显式边界约束，必要时仍需用户提供路径、描述或附件。拥有完整提示词的 profile 可以通过 Web 运行时的组合设置选择退出，而无需检查启动器路径。

.agents/notes/implemented/bug-fix/2026-07-28-web-gui-feedback-loop.i18n.yaml

@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-07-28-web-gui-feedback-loop.md
-2026-07-28-web-gui-feedback-loop.md: 039d2aebeeef903d10838a46b48e5172f0195126
-2026-07-28-web-gui-feedback-loop.zh.md: 34b6b26d4ce7c7e194e641536fffc503c013188b
+2026-07-28-web-gui-feedback-loop.md: aa488e1df087722c072d98c67d47cdf63a42f6b8
+2026-07-28-web-gui-feedback-loop.zh.md: d9de988e5a2cc57794ff628b65eb050dee610747

.agents/notes/implemented/bug-fix/2026-07-28-web-gui-feedback-loop.md

@@ -12,7 +12,7 @@ The [incident post-mortem](../../../../docs/postmortem/0003-web-agent-gui-feedba
 
 ## Decision
 
-`dsh web` publishes one canonical loopback URL and its actual runtime mode as both model-visible orientation and managed shell facts. The `app:web-surface` prompt section says that unqualified references identify this GUI and names the URL; `DSH_WEB_URL` and `DSH_WEB_MODE=production|development` carry the same facts into every foreground or managed background bash call. The section preserves the no-implicit-DOM, route, or screenshot boundary and does not claim that a LAN alias equals the browser's literal address.
+The ordinary `dsh web` composition mounts the Web bundle's `web-runtime` plugin, which publishes one canonical loopback URL and its actual runtime mode as both model-visible orientation and managed shell facts. The `app:web-surface` prompt section says that unqualified references identify this GUI and names the URL; `DSH_WEB_URL` and `DSH_WEB_MODE=production|development` carry the same facts into every foreground or managed background bash call. The section preserves the no-implicit-DOM, route, or screenshot boundary and does not claim that a LAN alias equals the browser's literal address. A complete-prompt profile sets the row's `surfaceContext` to false and receives neither the prompt section nor the managed variables; the Web launcher uses the same setting to suppress its source-checkout prompt section.
 
 The mode-specific prompt makes the agent, rather than the user, own the hidden startup contract. Production mode defines acceptance as rebuilding the affected artifacts and refreshing the existing URL. Development mode states that `dsh web --dev` activates only the HMR receiver: automatic client-plugin reload additionally requires a same-checkout `pnpm run dev:web` watcher, which the agent verifies before promising no-refresh updates. Shell and other plain-package changes still require rebuild plus refresh. An agent in production mode explains both commands when a user requests no-refresh updates; it does not launch a replacement GUI unless asked.
 
@@ -36,4 +36,4 @@ The keyless fresh-round-trip browser scenario boots the shipped production Web c
 
 ## Consequences
 
-Web prompts gain a dynamic URL-and-mode paragraph, so provider prefix reuse now varies by bound port and mode. Bash processes gain two non-secret managed environment variables. Bare Vite can no longer be used as a shell-only visual sandbox; developers use the full host or build mode instead. In exchange, GUI work has one mechanically observable target, the agent can teach the user the exact update behavior of the process actually serving their session, and the unsupported startup path fails before a white screen. The URL/mode contract guides the agent away from replacement ports; it does not prohibit arbitrary shell commands from starting one.
+Ordinary Web prompts gain a dynamic URL-and-mode paragraph, so provider prefix reuse now varies by bound port and mode. Their Bash processes gain two non-secret managed environment variables. Bare Vite can no longer be used as a shell-only visual sandbox; developers use the full host or build mode instead. In exchange, GUI work has one mechanically observable target, the agent can teach the user the exact update behavior of the process actually serving their session, and the unsupported startup path fails before a white screen. The URL/mode contract guides the agent away from replacement ports; it does not prohibit arbitrary shell commands from starting one. Profiles that disable `surfaceContext` also give up this feedback-loop guidance and shell context.

.agents/notes/implemented/bug-fix/2026-07-28-web-gui-feedback-loop.zh.md

@@ -12,7 +12,7 @@ Web agent（智能体）既无法识别承载当前会话的 GUI，也不知道
 
 ## 决策
 
-`dsh web` 发布一个规范的回环 URL 及其实际运行时模式，同时将二者作为模型可见的界面定位信息和受管 shell 事实。`app:web-surface` 提示词段说明：未加限定的指代指向此 GUI，并给出 URL；`DSH_WEB_URL` 和 `DSH_WEB_MODE=production|development` 会把同样的事实传入每次前台或受管后台 bash 调用。该段保留「不会隐式获得 DOM、路由或截图」这一边界，也不声称局域网别名等于浏览器中的实际地址。
+常规 `dsh web` 组合会挂载 Web 组合包的 `web-runtime` 插件，由它发布一个规范的回环 URL 及其实际运行时模式，同时将二者作为模型可见的界面定位信息和受管 shell 事实。`app:web-surface` 提示词段说明：未加限定的指代指向此 GUI，并给出 URL；`DSH_WEB_URL` 和 `DSH_WEB_MODE=production|development` 会把同样的事实传入每次前台或受管后台 bash 调用。该段保留「不会隐式获得 DOM、路由或截图」这一边界，也不声称局域网别名等于浏览器中的实际地址。拥有完整提示词的 profile 会把该配置行的 `surfaceContext` 设为 false，并且不会收到该提示词段和这些受管变量中的任何一个；Web 启动器也会使用同一项设置来抑制其源码 checkout 提示词段。
 
 按模式区分的提示词让 agent 而非用户负责隐藏的启动契约。生产模式将验收定义为重新构建受影响的产物并刷新现有 URL。开发模式说明，`dsh web --dev` 只会启用 HMR（热模块替换）接收端：客户端插件要自动重新加载，还需要在同一检出中运行 `pnpm run dev:web` 监听进程，agent 会在承诺无需刷新即可更新前验证这一点。外壳和其他普通包的变更仍然需要重新构建并刷新。生产模式下的 agent 会在用户要求无需刷新即可更新时说明这两个命令；除非用户要求，否则不会启动替代 GUI。
 
@@ -36,4 +36,4 @@ Web agent（智能体）既无法识别承载当前会话的 GUI，也不知道
 
 ## 影响
 
-Web 提示词会增加一个动态 URL 和模式段落，因此模型提供方的前缀复用会随绑定端口和模式变化。Bash 进程会增加两个非敏感的受管环境变量。裸 Vite 不再能用作只依赖 shell 的视觉沙箱；开发者应改用完整宿主或构建模式。作为交换，GUI 工作有了一个可由机制观察的唯一目标，agent 可以向用户说明实际承载其会话的进程究竟如何更新，不受支持的启动路径也会在出现白屏前失败。URL／模式契约会引导 agent 避免使用替代端口，但不会禁止任意 shell 命令启动替代服务。
+常规 Web 提示词会增加一个动态 URL 和模式段落，因此模型提供方的前缀复用会随绑定端口和模式变化。相应的 Bash 进程会增加两个非敏感的受管环境变量。裸 Vite 不再能用作只依赖 shell 的视觉沙箱；开发者应改用完整宿主或构建模式。作为交换，GUI 工作有了一个可由机制观察的唯一目标，agent 可以向用户说明实际承载其会话的进程究竟如何更新，不受支持的启动路径也会在出现白屏前失败。URL／模式契约会引导 agent 避免使用替代端口，但不会禁止任意 shell 命令启动替代服务。禁用 `surfaceContext` 的 profile 也会放弃这项反馈闭环指引与 shell 上下文。

.agents/notes/implemented/bug-fix/2026-07-29-sticky-composer-conversation-scroll.i18n.yaml

@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-07-29-sticky-composer-conversation-scroll.md
-2026-07-29-sticky-composer-conversation-scroll.md: d3fed7a9d0b1f39f9551fbd85e0f83515b1a2690
-2026-07-29-sticky-composer-conversation-scroll.zh.md: 2beee34d3bb68832d14b7607b43aa11e1425d53d
+2026-07-29-sticky-composer-conversation-scroll.md: 8657080078e475d6e90f17a6c0f8cb5b2fb1555d
+2026-07-29-sticky-composer-conversation-scroll.zh.md: a588903adf51c6fcdd01f851d4f5e6570f695242

.agents/notes/implemented/bug-fix/2026-07-29-sticky-composer-conversation-scroll.md

@@ -14,7 +14,7 @@ The active conversation column split scrolling: the chat (and trajectory) view o
 
 Session stats live on `'conversation.composer.dock'` (above `'conversation.input.dock'`). The InputBar textarea, when inside the host, chains `wheel` with `{ passive: false }`: while the capped textarea can still scroll in that direction it keeps the native gesture; only at its own edge does it `preventDefault` and apply `deltaY` to the host.
 
-Chat history prepend follows reader intent through stable rendered node/call identities rather than whole-scrollport height deltas. `ChatView` records the first visible `data-chat-anchor-key` and its top relative to the scrollport when paging starts, reselects the currently visible stable anchor after every reader scroll while the request is in flight, and compensates by that row's post-prepend rectangle delta. Reaching the bottom or appending the reader's own message cancels the paging anchor, so a late page cannot pull the view away from the newest content. Bottom follow is stored state rather than raw scroll geometry. A passive wheel listener takes its pre-input baseline from the last main-thread-delivered or programmatically written `scrollTop`, because Chromium may advance compositor geometry before delivering the event; the current non-negative floor excludes a concurrent layout clamp from reader movement. A scroll without matching wheel movement re-pins while following and only refreshes the semantic position while reading. ChatView's single `ResizeObserver` follows streaming, tool disclosure, and draft resize only while bottom ownership remains pinned, without a second per-chunk scroll write.
+Chat history prepend follows reader intent through stable rendered node/call identities rather than whole-scrollport height deltas. `ChatView` records the first visible `data-chat-anchor-key` and its top relative to the scrollport when paging starts, reselects the currently visible stable anchor after every reader scroll while the request is in flight, and compensates by that row's post-prepend rectangle delta. Reaching the bottom or appending the reader's own message cancels the paging anchor, so a late page cannot pull the view away from the newest content. Bottom follow is stored state rather than raw scroll geometry; how reader input is recognized — device-agnostic deviation from the observed-top ledger of the last delivered or written `scrollTop` — is owned by the [reader-scroll-attribution note](2026-08-06-reader-scroll-attribution-observed-top-ledger.md). ChatView's single `ResizeObserver` follows streaming, tool disclosure, and draft resize only while bottom ownership remains pinned, without a second per-chunk scroll write.
 
 ## Alternatives considered
 
@@ -26,8 +26,8 @@ Chat history prepend follows reader intent through stable rendered node/call ide
 
 **Keep StatsLine inside ChatView below the message column.** Rejected: outside the sticky composer it would scroll away while the input stayed pinned.
 
-**Model every browser scroll input source.** Rejected for this narrow fix: the reproduced desktop path uses wheel/trackpad input. Pointer/touch scrolling, native-scrollbar dragging, keyboard scrolling, focus navigation, and nested overflow ownership remain outside the provenance model instead of adding a general input state machine.
+**Model every browser scroll input source.** Rejected for this narrow fix: the reproduced desktop path uses wheel/trackpad input. Pointer/touch scrolling, native-scrollbar dragging, keyboard scrolling, focus navigation, and nested overflow ownership were left outside the provenance model instead of adding a general input state machine. The [reader-scroll-attribution note](2026-08-06-reader-scroll-attribution-observed-top-ledger.md) later closed this deferral by generalizing attribution through the observed-top ledger, still without an input state machine.
 
 ## Consequences
 
-Wheel over the footer scrolls the transcript; the visible layout is a fixed header, scrolling transcript, and sticky bottom composer. Stats appear on every active view tab. Nested view scrollers under the host are suppressed so sticky Turn headers in Trajectory stick to the column host. Concurrent history, streaming, tool expansion, and composer reflow preserve wheel/trackpad scroll decisions, including Chromium's compositor-first delivery and stream-finalization clamp/regrow. Other browser scroll inputs do not change follow ownership under this narrow provenance rule. No session → blank Hero and Hero → active both keep the same textarea DOM node and InputHub draft.
+Wheel over the footer scrolls the transcript; the visible layout is a fixed header, scrolling transcript, and sticky bottom composer. Stats appear on every active view tab. Nested view scrollers under the host are suppressed so sticky Turn headers in Trajectory stick to the column host. Concurrent history, streaming, tool expansion, and composer reflow preserve reader scroll decisions, including Chromium's compositor-first delivery and stream-finalization shrink clamp. Follow ownership extends to every reader input under the [reader-scroll-attribution note](2026-08-06-reader-scroll-attribution-observed-top-ledger.md). No session → blank Hero and Hero → active both keep the same textarea DOM node and InputHub draft.

.agents/notes/implemented/bug-fix/2026-07-29-sticky-composer-conversation-scroll.zh.md

@@ -14,7 +14,7 @@ Status: implemented
 
 会话统计挂在 `'conversation.composer.dock'`（位于 `'conversation.input.dock'` 之上）。InputBar 的 textarea 在宿主内以 `{ passive: false }` 链式处理 `wheel`：在限高 textarea 仍能沿该方向滚动时保留原生手势；仅在自身边缘才 `preventDefault` 并将 `deltaY` 施加到宿主。
 
-Chat 历史前插通过稳定的已渲染 node／call 身份跟随读者意图，而不是使用整个滚动容器的高度差。分页开始时，`ChatView` 记录第一个可见的 `data-chat-anchor-key` 及其相对滚动容器的顶部位置；请求在途期间，每次读者滚动都会重新选择当前可见的稳定锚点；页面到达后则按该行矩形的前后差值补偿。到达底部或追加读者自己的消息会取消分页锚点，因此迟到的页面不能把视图从最新内容拉走。贴底跟随采用存储状态，而不是原始滚动几何状态。passive wheel 监听器以最近一次由主线程交付或由程序写入的 `scrollTop` 作为输入前基线，因为 Chromium 可能先推进合成器几何状态，之后才交付事件；当前使用的非负下限不会将并发的布局钳制计入读者移动。没有对应滚轮／触控板输入位移的滚动，在跟随状态下会重新贴底，在阅读状态下则只刷新语义位置。`ChatView` 的单个 `ResizeObserver` 只会在贴底所有权仍保持时跟随流式输出、工具展开与草稿尺寸变化，且每个 chunk 不会触发第二次滚动写入。
+Chat 历史前插通过稳定的已渲染 node／call 身份跟随读者意图，而不是使用整个滚动容器的高度差。分页开始时，`ChatView` 记录第一个可见的 `data-chat-anchor-key` 及其相对滚动容器的顶部位置；请求在途期间，每次读者滚动都会重新选择当前可见的稳定锚点；页面到达后则按该行矩形的前后差值补偿。到达底部或追加读者自己的消息会取消分页锚点，因此迟到的页面不能把视图从最新内容拉走。贴底跟随采用存储状态，而不是原始滚动几何状态；读者输入如何被识别——即以与设备无关的方式偏离由最近一次交付或写入的 `scrollTop` 构成的 observed-top ledger——由[读者滚动归因笔记](2026-08-06-reader-scroll-attribution-observed-top-ledger.md)负责。`ChatView` 的单个 `ResizeObserver` 只会在贴底所有权仍保持时跟随流式输出、工具展开与草稿尺寸变化，且每个 chunk 不会触发第二次滚动写入。
 
 ## Alternatives considered
 
@@ -26,8 +26,8 @@ Chat 历史前插通过稳定的已渲染 node／call 身份跟随读者意图
 
 **把 StatsLine 留在 ChatView 消息列下方。** 否决：落在 sticky 编辑器之外会随内容滚走，而输入区仍钉在底部。
 
-**为每一种浏览器滚动输入来源建模。** 此次窄范围修复不采用：已复现的桌面端路径使用滚轮／触控板输入。指针／触控滚动、拖动原生滚动条、键盘滚动、焦点导航与嵌套 overflow 所有权仍不纳入输入来源模型，也不为此新增通用输入状态机。
+**为每一种浏览器滚动输入来源建模。** 此次窄范围修复不采用：已复现的桌面端路径使用滚轮／触控板输入。指针／触控滚动、拖动原生滚动条、键盘滚动、焦点导航与嵌套 overflow 所有权当时被留在输入来源模型之外，也未为此新增通用输入状态机。[读者滚动归因笔记](2026-08-06-reader-scroll-attribution-observed-top-ledger.md)后来通过 observed-top ledger 泛化了归因，补上了这一延后事项，且仍未引入输入状态机。
 
 ## Consequences
 
-在页脚上滚轮会滚动 transcript；可见布局是固定标题栏、可滚动 transcript 与 sticky 底部编辑器。统计出现在每一个活跃视图标签上。宿主下的嵌套视图 scroller 被抑制，因而 Trajectory 的 sticky Turn 标题贴在列宿主上。并发历史加载、流式输出、工具展开与编辑器重排会保留滚轮／触控板的滚动决定，包括 Chromium 先推进合成器几何状态再交付事件，以及流收尾阶段滚动位置受钳制后滚动容器重新增长的情况。在这条窄范围的输入来源规则下，其他浏览器滚动输入不会改变贴底跟随所有权。无 session → blank Hero 与 Hero → active 都保持同一 textarea DOM 节点以及 InputHub 草稿。
+在页脚上滚轮会滚动 transcript；可见布局是固定标题栏、可滚动 transcript 与 sticky 底部编辑器。统计出现在每一个活跃视图标签上。宿主下的嵌套视图 scroller 被抑制，因而 Trajectory 的 sticky Turn 标题贴在列宿主上。并发历史加载、流式输出、工具展开与编辑器重排会保留读者的滚动决定，包括 Chromium 先推进合成器几何状态再交付事件，以及流收尾阶段的收缩钳制。贴底跟随所有权依据[读者滚动归因笔记](2026-08-06-reader-scroll-attribution-observed-top-ledger.md)扩展到每一种读者输入。无 session → blank Hero 与 Hero → active 都保持同一 textarea DOM 节点以及 InputHub 草稿。

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.i18n.yaml

@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md
-2026-08-02-message-fork-actions-require-completed-turn-tail.md: f2e7fd67b65a6ce4a86ba3f4405f78842be8f234
-2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md: 2c3feeaa3ef01dbde67faa73257520918996f9c8
+2026-08-02-message-fork-actions-require-completed-turn-tail.md: abdcbc79948c67619bb70a8a87741046f65b8838
+2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md: a93b572c6db3c76ce3869747fd9a7660bf3ea395

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md

@@ -12,6 +12,8 @@ The Web conversation attached branch to the last assistant node with nonempty te
 
 `ConversationSnapshot.turnEnds` retains the completed turn boundaries present in the raw event window. The conversation view walks transcript nodes through each boundary and enables branch only when the boundary's last node is a user message, a durable steering message, or a content-bearing assistant message. Open turns have no eligible message, and a later tool result, reasoning-only interruption, turn error, or other transcript node leaves branch unavailable on earlier messages. The unavailable control stays visible, focusable, and hoverable; `aria-disabled`, a tooltip, and `aria-describedby` explain the completed-tail requirement without sending a Host request. Copy and clock remain available under their existing message chrome, and the Host's completed-turn fork semantics remain unchanged.
 
+The message-bubble half of this eligibility is superseded by the [user-bubble branch removal](../simplification/2026-08-06-user-bubbles-drop-the-branch-action.md): user and steering bubbles no longer render the control at all, so only content-assistant tails may fork; the assistant-side gate and its visible-but-unavailable presentation stand.
+
 This narrows the message eligibility established by the earlier [Web session fork action decision](../feature/2026-07-27-web-session-fork-actions.md). Session-row forking still selects the latest completed turn, and eligible message actions still pass their event seq through the shared client runtime operation.
 
 ## Alternatives considered

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md

@@ -12,6 +12,8 @@ Web 会话把分支操作挂到每个轮次中最后一个文本非空的 assist
 
 `ConversationSnapshot.turnEnds` 保留原始事件窗口中的已完成轮次边界。会话视图按各边界遍历 transcript（文本记录）节点，仅当边界的最后一个节点是用户消息、持久 steering（中途引导）消息或含内容的 assistant 消息时才启用分支操作。开放轮次没有符合条件的消息；如果后面还有工具结果、只有推理内容的中断、轮次错误或其他 transcript 节点，较早消息上的分支操作会保持不可用。不可用的控件仍然可见、可聚焦、可悬停；`aria-disabled`、tooltip 与 `aria-describedby` 会说明已完成尾部这一要求，且不会发送 Host 请求。复制和时钟仍可在既有消息 chrome 下使用，Host 按已完成轮次 fork 的语义保持不变。
 
+本资格判定中消息气泡的那一半已被 [user 气泡分支移除决策](../simplification/2026-08-06-user-bubbles-drop-the-branch-action.md)取代：user 与 steering 气泡不再渲染该控件，因此只有内容 assistant 尾部可以 fork；assistant 侧门禁及其可见但不可用的呈现保持有效。
+
 本决策收紧了较早的 [Web 会话 fork 操作决策](../feature/2026-07-27-web-session-fork-actions.md)所定义的消息资格。Session 行 fork 仍选择最新的已完成轮次；符合条件的消息操作仍通过共享 client 运行时操作传递其事件 seq。
 
 ## 考虑过的替代方案

.agents/notes/implemented/bug-fix/2026-08-04-conversation-column-one-axis-scroll.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/2026-08-04-conversation-column-one-axis-scroll.md
+2026-08-04-conversation-column-one-axis-scroll.md: 9a487c506a75033d0854f08e95da24704309003d
+2026-08-04-conversation-column-one-axis-scroll.zh.md: 23441a7c8655d1f19d3c0fe0f661f81f69b55dba

.agents/notes/implemented/bug-fix/2026-08-04-conversation-column-one-axis-scroll.md

@@ -0,0 +1,37 @@
+# Agent Note: The conversation column scrolls on one axis
+
+Status: implemented
+
+English | [中文](2026-08-04-conversation-column-one-axis-scroll.zh.md)
+
+## Problem
+
+Narrowing the center column — by the window or by the sidebar drag — put a horizontal scrollbar under the whole conversation column on the hero. The bleeding element is the hero's decorative backdrop ellipse: `.heroGlow` is sized `1051/776` of the hero box so its blur scales in userSpace with the input card, which means it reaches past the column whenever the column is narrower than the glow.
+
+That bleed is by construction and stays. What made it user-visible is the scroll container it sits in. `[data-conversation-scroll]` declared `overflow-y: auto` and left the other axis at its initial `visible`, and a box that scrolls in one axis computes `visible` to `auto` in the other. Every column narrower than the glow therefore offered a real horizontal scroll range — measured at 24–95px across the widths a laptop actually produces.
+
+## Decision
+
+`.scrollBody` declares `overflow-x: hidden`. The column states that it is a one-axis scroller instead of leaving the second axis to be derived.
+
+Clipping does not change. `overflow-y: auto` had already made the box a scroll container that clips both axes, so the declaration withdraws only the scrollbar and the user gesture; the glow keeps its bleed, its blur radius, and the same painted extent, and the column keeps its vertical scroll. Nothing in the composer chain moves.
+
+## Alternatives considered
+
+**Size the glow to fit the column.** Rejected. The glow's width is what scales its `stdDeviation="50"` blur with the input card (figma 313:14109); constraining it would make the blur tighten as the column narrows, which is a visual regression to fix a scrollbar.
+
+**Wrap the glow in a clipping box.** Rejected. It adds a box whose only job is to undo an overflow the column already clips, and it leaves the derived `overflow-x: auto` in place for the next element that bleeds — the transcript is full of candidates.
+
+**Rely on the frame's `.centerCol { overflow: hidden }`.** It cannot help. That clip is outside the scroll container, so it hides the glow's overhang at the column border while the container inside it still scrolls to reach it. The reported bar was that container's.
+
+**Assert `scrollWidth === clientWidth` in the test.** Rejected as the signal, because it does not distinguish the states: `hidden` clips the bleed rather than reflowing it away, so the scroll range reads the same on both sides of the fix. Only refusing a user gesture differs, which is what the scenario measures.
+
+## Testing
+
+[apps/web/tests/conversation-column-overflow.e2e.ts](../../../../apps/web/tests/conversation-column-overflow.e2e.ts) sweeps viewport widths bracketing the glow and, at each stop, wheels horizontally over the column and reads `scrollLeft`. The committed golden records the relation per stop; the widest stop is the control where the glow does not bleed at all.
+
+Two guards keep the scenario honest. The vacuity guard asserts the glow still reaches past the column at the narrow stops, so the claim cannot pass by the symptom having disappeared for an unrelated reason. The mutation control forces `overflow-x: auto` back on in the page and shows the same gesture, at the same timing, carrying the column to its positive scroll boundary; the test measures that boundary directly because a stable scrollbar gutter can leave some overflow on the negative side of the scroll origin. Without the control, a `scrollLeft` of 0 could equally mean the wheel never arrived.
+
+## Consequences
+
+The conversation column no longer offers a horizontal scrollbar at any width, and decorative bleed in the composer chain is now clipped rather than exposed as scroll range. The cost is that genuinely wide content under this column is clipped instead of reachable by scrolling: any such surface owns its own scroller, as the markdown code block and the trajectory table already do.

.agents/notes/implemented/bug-fix/2026-08-04-conversation-column-one-axis-scroll.zh.md

@@ -0,0 +1,37 @@
+# Agent Note：会话列只在一个轴上滚动
+
+状态：已实现
+
+[English](2026-08-04-conversation-column-one-axis-scroll.md) | 中文
+
+## 问题
+
+当中间列被拉窄——无论是拖窗口还是拖侧边栏——hero 态的整条会话列下方就会出现一条横向滚动条。溢出的元素是 hero 的装饰性背景椭圆：`.heroGlow` 的宽度取 hero 盒子的 `1051/776`，好让它的模糊在 userSpace 中随输入卡片一同缩放；这也意味着只要列比它窄，它就会伸出列外。
+
+这处外溢是设计使然，保持不变。真正让它对用户可见的是它所处的滚动容器。`[data-conversation-scroll]` 只声明了 `overflow-y: auto`，另一个轴留在初始值 `visible`；而一个在某一轴上滚动的盒子，会把另一轴的 `visible` 计算为 `auto`。于是每一条比该椭圆窄的列都真的给出了一段横向滚动范围——在笔记本实际会产生的几档宽度上，实测为 24–95px。
+
+## 决定
+
+`.scrollBody` 声明 `overflow-x: hidden`。这条列明确声明自己是单轴滚动容器，而不是把第二个轴交给推导。
+
+裁剪行为不变。`overflow-y: auto` 早已使该盒子成为在两个轴上都裁剪的滚动容器，因此这条声明收回的只是滚动条和用户手势；椭圆保留它的外溢、模糊半径和同样的绘制范围，列也保留纵向滚动。输入区那条链路上没有任何东西移动。
+
+## 曾考虑的替代方案
+
+**把椭圆缩到列内。** 否决。椭圆的宽度正是让它 `stdDeviation="50"` 的模糊随输入卡片缩放的依据（figma 313:14109）；约束宽度会使列越窄模糊越紧，等于为修一条滚动条而制造一处视觉回归。
+
+**给椭圆套一层裁剪盒。** 否决。这层盒子唯一的职责是抵消列本就会裁剪的溢出，而推导出的 `overflow-x: auto` 仍然留在原处，等着下一个外溢的元素——会话流里这样的候选者不少。
+
+**依赖外框的 `.centerCol { overflow: hidden }`。** 它帮不上忙。那处裁剪在滚动容器之外，只能在列边界处遮住椭圆探出的部分，而里面的容器照样可以滚过去够到它。用户报告的那条滚动条属于内层容器。
+
+**在测试里断言 `scrollWidth === clientWidth`。** 作为判据被否决，因为它区分不出两种状态：`hidden` 裁剪外溢，而不是把它重排掉，所以修复前后读到的滚动范围一样。唯一有差别的是拒绝用户手势，这正是该场景所测量的。
+
+## 测试
+
+[apps/web/tests/conversation-column-overflow.e2e.ts](../../../../apps/web/tests/conversation-column-overflow.e2e.ts) 扫过一组把椭圆宽度夹在中间的视口宽度，在每一档上向列横向滚轮并读取 `scrollLeft`。提交的 golden 逐档记录该关系；最宽的一档是椭圆根本不外溢的对照。
+
+两道防线保证该场景不流于形式。空断言防线断言窄档上椭圆确实仍伸出列外，使这项主张不可能因为症状出于无关原因消失而通过。变异对照则在页面内把 `overflow-x: auto` 强制改回，证明同一手势在同一时序下能把列带到正向滚动边界。测试直接测量该边界，因为稳定的滚动条槽可能让部分外溢处于滚动原点的负向。没有这项对照，`scrollLeft` 读到 0 同样可以解释为滚轮根本没送达。
+
+## 后果
+
+会话列在任何宽度下都不再给出横向滚动条，输入区链路上的装饰性外溢从暴露为滚动范围改为被裁剪。代价是这条列下真正过宽的内容会被裁掉而非可滚动够到：这类界面各自拥有自己的滚动容器，markdown 代码块和轨迹表格已经如此。

.agents/notes/implemented/bug-fix/2026-08-05-turn-tail-actions-require-a-completed-turn.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/2026-08-05-turn-tail-actions-require-a-completed-turn.md
+2026-08-05-turn-tail-actions-require-a-completed-turn.md: 689d50bb86c830d6e428239f112568f00d74c9b8
+2026-08-05-turn-tail-actions-require-a-completed-turn.zh.md: 2cc426bbb82acb8f57d491b0f068e89771699357

.agents/notes/implemented/bug-fix/2026-08-05-turn-tail-actions-require-a-completed-turn.md

@@ -0,0 +1,31 @@
+# Agent Note: Turn-tail IconActions require a completed turn
+
+Status: implemented
+
+English | [中文](2026-08-05-turn-tail-actions-require-a-completed-turn.zh.md)
+
+## Problem
+
+Assistant IconActions were derived from the finalized transcript alone: the last content-text assistant of each turn owned the row. That quantity is stable only after the turn closes. While a turn is still producing steps, the narration a model writes before a tool call *is* the last content assistant so far, so it took the row for as long as the tool ran and then lost it to the next step's text. Readers saw copy, branch, and a clock appear under an intermediate sentence, shift the flow by one 28px row, and disappear. The row was also incoherent in that state: its branch control was already disabled through `turnEnds`, and its `Ran for` label was already withheld through `turnTimings`, so only copy worked.
+
+The [message chrome decision](../feature/2026-07-29-web-message-icon-actions-and-clock.md) always claimed mid-turn narration stays chrome-free; the derivation never carried a completion signal to make that true.
+
+## Decision
+
+`assistantActionsSeqs` takes `ConversationSnapshot.turnEnds` and grants the row only within a turn that has a `turn/end` in the window. Ownership inside a completed turn is unchanged: its last content-text assistant. A turn still producing steps grants nothing, so its narration never mounts the row, and the seat appears once, under the settled answer, when the turn closes.
+
+This is the same completion fact the branch control and the run-time label already use, so the three parts of one row now agree. Turn completion is read from the durable `turn/end` event rather than inferred from `running`, the streaming partial, or in-flight tool calls, matching the [completed-turn-tail decision](2026-08-02-message-fork-actions-require-completed-turn-tail.md). Every reason kind closes a turn, so an aborted turn's frozen tail keeps its footer, and a crash-orphaned turn receives its `turn/end` from log repair on load.
+
+`hasContentText` moves to `chat-flow.ts` and `AssistantMarkdown` imports it, so the ownership gate and the mount gate cannot drift apart.
+
+## Alternatives considered
+
+**Withhold by naming the open turn from `running` plus the streaming partial or the first in-flight tool call.** This shipped briefly in the original change and was then dropped. It infers completion instead of reading it, needs a special case so a turn accepted before its first step does not strip the previous answer's seat, and is the inference the completed-turn-tail decision rejected for the branch control. `turnEnds` answers the same question per turn with no inference and no special case.
+
+**Leave the row mounted mid-turn and disable its controls.** Rejected: mid-turn narration is not a degraded answer, it is not the answer. Copy would still write an intermediate sentence, and the row would still move to the real tail at turn end.
+
+**Keep the row under every finalized content node permanently.** Rejected again here for the reason the original decision gave: repeating copy, branch, and a clock under every step clutters the flow. It also does not solve the reported problem, since the branch control is only meaningful on the tail.
+
+## Consequences
+
+A running turn carries no message footer below the user bubble that triggered it, while every earlier completed turn keeps its own; the seat appears once when `turn/end` lands, which adds one 28px row under the settled answer at that moment. A turn whose `turn/end` is outside the loaded window grants nothing, which cannot arise from paging because a turn's end follows its own nodes. `apps/web/tests/turn-tail-actions.e2e.ts` pins both states through the assembled application: a `hang` sidecar on the second model call parks a turn whose first step narrated before calling bash, and the two goldens hold the parked flow and the flow after stopping. Package tests cover the derivation directly and the running-turn render.

.agents/notes/implemented/bug-fix/2026-08-05-turn-tail-actions-require-a-completed-turn.zh.md

@@ -0,0 +1,31 @@
+# Agent Note: 轮次尾部 IconActions 要求轮次已完成
+
+Status: implemented
+
+[English](2026-08-05-turn-tail-actions-require-a-completed-turn.md) | 中文
+
+## 问题
+
+assistant IconActions 此前只从已定稿的 transcript（文本记录）推导：每个轮次中最后一条含内容文本的 assistant 拥有该行。这个量只有在轮次关闭后才稳定。轮次仍在产出步骤时，模型在工具调用前写下的叙述就是当时该轮次的最后一条内容 assistant，于是它在工具执行期间取得该行，等下一步的文本落定又把它交出去。读者会看到复制、分支和时钟出现在一句中间叙述下方，把流程推开一行 28px，然后消失。该行在这个状态下本身也是残缺的：分支控件已经通过 `turnEnds` 判定为禁用，`Ran for` 标签已经通过 `turnTimings` 判定为不显示，只有复制可用。
+
+[消息 chrome 决策](../feature/2026-07-29-web-message-icon-actions-and-clock.md)一直声称轮次中间的叙述不带 chrome，但推导过程从未拿到能让这句话成立的完成信号。
+
+## 决策
+
+`assistantActionsSeqs` 接收 `ConversationSnapshot.turnEnds`，只在事件窗口中存在该轮次 `turn/end` 时才授予该行。已完成轮次内部的归属不变，仍是其最后一条含内容文本的 assistant。仍在产出步骤的轮次不授予任何座位，因此其叙述不会挂载该行；轮次关闭时，座位在已定稿答案下方一次性出现。
+
+这与分支控件和运行时长标签使用的完成事实相同，因此同一行的三个部分现在口径一致。轮次是否完成读自持久的 `turn/end` 事件，而不是从 `running`、流式 partial 或在途工具调用推断，与[已完成轮次尾部决策](2026-08-02-message-fork-actions-require-completed-turn-tail.md)一致。任何 reason 类别都会关闭轮次，因此已中止轮次冻结的尾部保留其操作栏，而崩溃遗留的开放轮次会在加载时由日志修复补上 `turn/end`。
+
+`hasContentText` 移入 `chat-flow.ts`，由 `AssistantMarkdown` 导入，使归属门控与挂载门控无法各自漂移。
+
+## 考虑过的替代方案
+
+**用 `running` 加流式 partial 或第一个在途工具调用指认开放轮次，据此扣留。** 这一做法曾在最初的变更中短暂存在，随后被删除。它推断完成状态而不是读取完成状态，还需要一个特例，避免轮次已被接受但尚未产出第一步时把上一条回答的座位取走；这正是已完成轮次尾部决策为分支控件否决过的推断。`turnEnds` 按轮次回答同一个问题，不需要推断，也不需要特例。
+
+**轮次进行中保留该行，只把控件置为不可用。** 不予采纳：轮次中间的叙述不是一个降级的答案，它根本不是答案。复制仍然会写入一句中间文本，该行在轮次结束时仍然要移动到真正的尾部。
+
+**让每个已定稿的内容节点长期保留该行。** 在此重新否决，理由与最初的决策相同：在每一步下重复复制、分支和时钟会打乱流程。它也解决不了本次报告的问题，因为分支控件只有落在尾部才有意义。
+
+## 后果
+
+运行中的轮次在触发它的用户气泡之下不再有任何消息操作栏，而此前每个已完成轮次仍保留各自的座位；座位在 `turn/end` 到达时一次性出现，此刻已定稿答案下方会多出一行 28px。`turn/end` 落在加载窗口之外的轮次不授予座位，而翻页不会造成这种情况，因为一个轮次的结束事件排在它自己的节点之后。`apps/web/tests/turn-tail-actions.e2e.ts` 通过组装后的应用钉住两种状态：`hang` sidecar 作用在第二次模型调用上，把一个首步先叙述再调用 bash 的轮次挂住，两份 golden 分别记录挂起中的流程和停止之后的流程。包级测试直接覆盖该推导以及运行中轮次的渲染结果。

.agents/notes/implemented/bug-fix/2026-08-06-api-key-format-validation.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/2026-08-06-api-key-format-validation.md
+2026-08-06-api-key-format-validation.md: d1f6d31d362b76392514704be780f553b45d36ad
+2026-08-06-api-key-format-validation.zh.md: 75b3fa247bdf449964a874e909e6e3bc9e0694fa

.agents/notes/implemented/bug-fix/2026-08-06-api-key-format-validation.md

@@ -0,0 +1,107 @@
+# Agent Note: Validate API key format before it reaches an HTTP header
+
+Status: implemented
+
+English | [中文](2026-08-06-api-key-format-validation.zh.md)
+
+## Problem
+
+An API key holding characters no HTTP header value can carry was accepted by every configuration surface and failed only when a request was built, far from the field that caused it.
+
+Pasting a key containing an emoji, CJK text, or a full-width punctuation mark into the web Models page reported a successful save. The first turn then failed with `Cannot convert argument to a ByteString because the character at index 7 has a value of 55357 which is greater than 255` — the index and code point are UTF-16 internals with no action attached, and they disclose the code point of one character of the key. `llm-deepseek` produced this because `fetch` builds the `Bearer` header inside the `try` in [adapter.ts](../../../../packages/llm/llm-deepseek/src/adapter.ts), whose `catch` labels every failure `TRANSPORT`; that label is in `DEFAULT_RETRYABLE_CODES`, so a permanent, deterministic fault was also retried three times.
+
+`llm-pi-ai` was worse on the same input. Its discovery probe builds the same header with a bare `fetch` in [discovery.ts](../../../../packages/llm/llm-pi-ai/src/discovery.ts) and wrapped every failure as `could not reach <url>`, so a local key fault was reported as an unreachable network. The probe is reachable from the unsaved draft: `ProviderEditor` puts the typed `keyDraft` into its probe request, so the model-listing button sent an illegal key before anything was stored.
+
+Whitespace passed every check. `ProviderEditor` tested `keyDraft.length` and `resolveAdapterOptions` tested `config.apiKey.length`, so a key of three spaces stored and then authenticated as `Bearer` plus blanks. `llm-pi-ai` rejected an empty literal `apiKey` in `resolveProfiles`, but applied no check whatsoever to a credential- or environment-sourced key — the path the Models page writes, and therefore the path users actually take.
+
+Sources: deepseek-harness#1594 and #1595; dsh-external#247, #249, #266, and #210.
+
+## Decision
+
+One rule defines a legal key: **after trimming, non-empty, and every character within `[\x21-\x7E]`** — printable ASCII, space excluded.
+
+This single predicate covers every input the sources list: empty, leading and trailing whitespace, interior whitespace, C0 control characters, emoji, CJK text, and full-width punctuation. It is also exactly the constraint that produced the ByteString failure, so the two issues close on one definition rather than on two coincidentally related fixes.
+
+A second, narrower rule catches a pasted environment line: input matching `^[A-Z][A-Z0-9_]*=[^=]` or wrapped in matching quotes is refused. Restricting the prefix to upper-case keeps real keys clear of it — `sk-` forms break the identifier match at the hyphen — and requiring a non-`=` character after the separator keeps base64 padding clear of it too. It reports the same format failure as an illegal character rather than its own message: the reader's next move is identical either way, so a separate line would name a cause without changing what to do.
+
+### Invariants belong at every layer; heuristics belong where the human is
+
+The charset rule is an invariant. A non-ASCII character *cannot* travel in a header value for any provider, so enforcing it in the browser, in each resolver, and on every credential read is consistent by construction rather than by agreement.
+
+The shape rule is a guess about how people paste, so it runs **only in the browser**. `llm-pi-ai` fronts OpenAI, Anthropic, and arbitrary hand-declared gateways whose key formats this repository does not own; a gateway issuing a key shaped like `TENANT1=abc` would, if the rule ran in the resolver, be locked out with no escape — the settings page would refuse it and a hand-written `.env` would be rejected on read. Confining the heuristic to the surface where the paste happens keeps the environment as the way through.
+
+### Absence is a configuration state, not a missing key
+
+"No API key" means three different things here, and only one of them is an error. The rule applies to a value that was *provided*; deciding whether one was provided at all stays with each caller.
+
+**Omitted.** A profile naming neither `apiKey` nor `apiKeyEnv` is authenticated by something other than a harness-held key. `routeAuth` in [provider.ts](../../../../packages/llm/llm-pi-ai/src/provider.ts) keeps the installed catalog provider's own auth precisely so provider-native ambient discovery survives, and `openai-codex` — shipped in that catalog — authenticates through OAuth and refuses an explicit key outright. `namesCredential` carries this distinction. In `llm-deepseek`, an absent `apiKey` likewise falls through to `apiKeyEnv`. Omission is never validated.
+
+**A blank field in the web UI.** The key input opens empty even for a provider whose key is already stored — the `keyStored` copy reads "Configured — enter a new value to replace" — so blank means *keep what is stored*. `ProviderEditor` skips `credentials.set` entirely when the draft is empty, and that stays a no-op: a blank field never blocks submit, or editing a base URL would demand re-entering the key.
+
+**Provided, but empty or whitespace-only.** What this means depends on what absence selects for that surface, and the two adapters differ for a reason. In `llm-pi-ai` it is an error, because absence there switches authentication mode — to the installed provider's ambient discovery or OAuth — so a blank key leaves genuine ambiguity about which was meant; its wording names the legitimate alternative rather than just refusing (*has an empty apiKey; omit it to use ambient authentication*). In `llm-deepseek` absence merely selects a different *source* for the same key, `apiKeyEnv`, so a blank literal resolves through that fallback exactly as an omitted one does. In the browser it is always a failure, on both cards: the field is where a person just typed, and silently discarding what they typed is never the right answer.
+
+`normalizeApiKey` therefore takes `string`, never `string | undefined`.
+
+### Where the rule lives
+
+`normalizeApiKey` is a module of the `dsh-llm` seam, beside [attribution.ts](../../../../packages/llm/llm/src/attribution.ts), which already owns shared header concerns. Both adapters depend on the seam and both need the rule, so it has two current consumers rather than a speculative one. It returns the trimmed value or a reason (`empty`, `illegalCharacters`).
+
+Both adapters also need the identical "refuse a stored credential" diagnosis, differing only by package prefix. `LlmError` is declared in the seam's `index.ts`, so `assertUsableApiKey(raw, pkg, ref)` lives there beside it and neither adapter carries a local copy. The predicate module stays dependency-free: importing `LlmError` into `api-key.ts` would cycle with `index.ts`'s re-export of it.
+
+The client cannot import any of this: client packages reference only client packages, so `packages/client/ui-models` mirrors the predicate in its own `apiKey.ts` and owns the localized messages, exactly as `validateDeepSeekModels` mirrors the host's `catalogModel` schema. Each side names the other in a comment.
+
+### What each surface does
+
+| Surface | Behavior |
+|---|---|
+| `dsh-llm` | Owns `normalizeApiKey`, `assertUsableApiKey`, and `INVALID_CREDENTIAL_CODE`, which is deliberately outside `DEFAULT_RETRYABLE_CODES`. |
+| `llm-deepseek` `resolveAdapterOptions` | Refuses a literal `apiKey` no header can carry, beside the other beyond-schema bounds; uses the trimmed value. An absent or blank one falls through to `apiKeyEnv`. |
+| `llm-deepseek` `resolveApiKey` | Normalizes what the credentials seam or environment returns, rejecting with `INVALID_CREDENTIAL` naming the Models page and never echoing the key. |
+| `llm-pi-ai` `resolveProfiles` | Applies the shared rule, keeping its "omit it to use ambient authentication" wording, and writes the trimmed value into the resolved profile. |
+| `llm-pi-ai` `resolveApiKey` | Normalizes the credential and environment paths. A profile naming no credential still returns `undefined`, so ambient and OAuth routes are unaffected. |
+| `llm-pi-ai` `discoverModels` | Normalizes before building the header, so an illegal key is a credential fault rather than an unreachable endpoint. A probe carrying no key stays unauthenticated. |
+| `ui-models` | Mirrors the charset rule, adds the shape heuristic, trims `keyDraft` before probe and `credentials.set`, and fixes the `stringAt` emptiness test. A blank field remains a no-op that submits; a field holding only whitespace is a field-level failure. Submit **and the endpoint interrogation** are both gated, so a refused key never spends a round trip to be told what the field already says, and the failure renders on the field, matching the existing `modelFailure` pattern. |
+
+`ProviderEditor` serves both the DeepSeek and pi-ai layouts, so one client change covers both providers. `CustomProviderCard` carries the same judgement for a hand-declared route.
+
+`credentials-local` is deliberately untouched. It stores credentials generally, and printable-ASCII is a constraint of HTTP headers rather than of credential storage; its existing refusal of values no dotenv style can represent stands as it was.
+
+## Alternatives considered
+
+**A `.pattern()` on the `apiKey` schema field.** Vendored schemastery supports it, and the pattern would serialize to the browser with the rest of the namespace schema — one rule, delivered rather than mirrored. It lost because a pattern cannot trim first: `cordis.yml` would then reject a padded key while `.env` tolerated one, and the resolver would disagree with the schema about the same string. Validating in `resolveAdapterOptions` keeps every surface trim-then-validate, and that function is already where this package re-judges bounds the schema cannot express.
+
+**A validation module shared by client and host.** Rejected by the source-plane layout: client packages reference only client packages plus `vendor/cordis` and `support/invariants`, and widening that to reach a host package would collide the two `Context` merges the split exists to keep apart. Mirroring a one-line predicate with a test on each side is the established shape here.
+
+**A per-adapter thrower in each of `llm-deepseek` and `llm-pi-ai`.** The first plan gave each adapter its own, differing only by the package prefix in the message, with a duplication-gate exemption to excuse the pair. Rejected before implementation: `LlmError` is declared in the seam, so the seam can own the diagnosis outright, and an exemption there would have hidden exactly the duplication it was covering for.
+
+**Sniffing the `TypeError` in the adapter's `catch`.** This would classify the ByteString failure after the fact, leaving the header construction itself unguarded. It depends on the wording of a Node error message, so it degrades silently across runtime versions, and it cannot help `llm-pi-ai`, whose request header is built inside the pi-ai SDK. Refusing the key before handing it over works for both adapters and for the discovery probe.
+
+**Enforcing in `credentials-local.set`.** It would catch every writer at once, including a hand-edited file. It lost because that provider stores credentials of every kind, and a rule derived from HTTP header encoding does not belong to it.
+
+**Running the shape heuristic in the resolvers too.** Symmetric, and it would stop a pasted environment line written directly into `.env`. Rejected for the lockout described above: a false positive in a resolver leaves the user no working path, while a false positive in the browser leaves the environment open.
+
+**Probing the provider at save time to prove the key works.** It would close the complaint the sources actually open with — a save that reports success and fails at the first turn. Rejected as out of scope and, on the code as it stood, unbuildable: `discoverModels` short-circuits to the installed catalog before any network call for exactly the providers pi-ai ships catalogs for, so it verified nothing about the key, and the DeepSeek card has no probe at all. A verifier's value is distinguishing "key rejected" from "cannot reach", which is the distinction this change makes reliable; building it first would have produced a verifier unable to tell its own outcomes apart. Comparable products also do not verify on save, so a blocking network call there would be an unexpected behavior rather than a missing one.
+
+## Consequences
+
+A malformed key is refused at the field that holds it, and a malformed stored key fails as `INVALID_CREDENTIAL` with a message naming where to fix it and no fragment of the key. Because that code sits outside `DEFAULT_RETRYABLE_CODES`, a deterministic credential fault is no longer retried three times as a transport blip. `llm-pi-ai` discovery reports an illegal probe key as a credential fault instead of an unreachable endpoint.
+
+The shape heuristic can refuse a real key. The first draft matched any upper-case identifier followed by `=`, which review showed was broader than intended: an all-upper-case base64 key ending in padding (`ABCD==`) matched an assignment it does not resemble. Requiring a non-`=` character after the separator excludes padding, since base64 only ever pads at the end. What remains — an upper-case name, one `=`, then a value — is a shape no known provider issues, and the rule runs only in the browser, so a user who still hits it can set the credential through the environment. The residual cost is a confusing refusal for a key nobody has yet reported.
+
+Restricting to printable ASCII is stricter than the transport requires: a header value may carry `\x80`–`\xFF`. Admitting latin-1 would let `é` through to return an opaque 401 instead of a local, explained refusal, so the stricter rule is deliberate. A provider that issues latin-1 keys would need this rule widened.
+
+The charset predicate exists twice, once per source plane. The layout forbids sharing it; each side carries its own test and names its twin.
+
+Keys already stored by an earlier build are read through `resolveApiKey`, so an illegal stored value fails at resolution rather than at request time. The diagnosis improves, but the failure moves earlier for anyone currently holding one.
+
+The costliest way to get this wrong would have been to treat absence as invalidity: a rule applied to `undefined` breaks every route authenticating through ambient discovery or OAuth, and a blank field that blocked submit makes editing any other setting demand re-entering the key. Both are pinned by tests rather than left to care.
+
+## Testing
+
+`packages/llm/llm/tests/api-key.spec.ts` drives `normalizeApiKey` and `assertUsableApiKey` over the whole input table — empty, whitespace-only, padded, interior-space, C0 control, emoji, CJK, full-width, latin-1, and the printable-ASCII boundary — and pins that a refusal carries `INVALID_CREDENTIAL` and no part of the key.
+
+`packages/llm/llm-deepseek/tests/` covers the literal-config path in `adapter.spec.ts` and the stored-credential path end to end in `dynamic-config.spec.ts`, through the real credentials seam rather than a stub. `packages/llm/llm-pi-ai/tests/` covers `resolveProfiles` — including that the trimmed value reaches the resolved profile, which the `...rest` spread would otherwise discard — and the discovery probe, including that a probe with no key sends no `authorization` header.
+
+`packages/client/ui-models/tests/` pins `apiKeyFailure` over the same table plus the paste-shape cases, and drives both cards: a blank field submits without writing a credential, a whitespace-only field fails on the field, an illegal or wrapped key blocks submit and the interrogation alike, a padded key is trimmed before `credentials.set` and before an interrogation, and a hand-declared route can be created with no key at all.
+
+The user-visible terminal state is pinned where it is actually assembled: `examples/headless-agent/tests/headless.snapshot.ts` runs the one-shot app against a stored key no header can carry, over the same keyless composition its missing-credential sibling uses, and records that the turn ends on `INVALID_CREDENTIAL` with an actionable message carrying neither the key nor the word `ByteString`. A package test could not have shown that, and the web e2e covers only the browser half.

.agents/notes/implemented/bug-fix/2026-08-06-api-key-format-validation.zh.md

@@ -0,0 +1,107 @@
+# Agent Note: 在 API Key 进入 HTTP header 之前校验其格式
+
+Status: implemented
+
+[English](2026-08-06-api-key-format-validation.md) | 中文
+
+## Problem
+
+一个含有 HTTP header value 无法承载的字符的 API Key，曾被每一层配置界面接受，直到构造请求时才失败——离引发它的那个字段已经很远。
+
+把含 emoji、中文或全角标点的 Key 粘进 Web 模型设置页，保存会报成功。第一轮对话随即失败于 `Cannot convert argument to a ByteString because the character at index 7 has a value of 55357 which is greater than 255`——其中的下标与码点是 UTF-16 内部细节，不附带任何可执行动作，却泄露了 Key 中某一个字符的码点。`llm-deepseek` 之所以产出这句，是因为 `fetch` 在 [adapter.ts](../../../../packages/llm/llm-deepseek/src/adapter.ts) 的 `try` 内部构造 `Bearer` header，而那个 `catch` 把一切失败都标为 `TRANSPORT`；该标签又在 `DEFAULT_RETRYABLE_CODES` 之中，于是一个永久且确定的故障还会被重试三次。
+
+同样的输入在 `llm-pi-ai` 上更糟。它的探测路径在 [discovery.ts](../../../../packages/llm/llm-pi-ai/src/discovery.ts) 里用裸 `fetch` 构造同一个 header，并把一切失败包装成 `could not reach <url>`，于是一个本地的 Key 故障被报成网络不可达。这条探测在保存之前就够得着：`ProviderEditor` 把用户输入的 `keyDraft` 直接放进探测请求，所以「获取模型列表」按钮会在任何东西落盘之前就把非法 Key 发出去。
+
+空白字符能通过每一道检查。`ProviderEditor` 判的是 `keyDraft.length`，`resolveAdapterOptions` 判的是 `config.apiKey.length`，于是三个空格构成的 Key 会被存下，随后以 `Bearer` 加若干空格去认证。`llm-pi-ai` 在 `resolveProfiles` 中拒绝空的字面量 `apiKey`，却对来自凭据或环境的 Key 完全不做检查——而那正是模型设置页写入的路径，也就是用户真正走的路径。
+
+来源：deepseek-harness#1594 与 #1595；dsh-external#247、#249、#266、#210。
+
+## Decision
+
+一条规则定义什么是合法 Key：**trim 之后非空，且每个字符都落在 `[\x21-\x7E]`**——可打印 ASCII，不含空格。
+
+这一个断言覆盖了来源列出的全部输入：空值、首尾空白、中间空白、C0 控制字符、emoji、中文、全角标点。它同时正是造成 ByteString 失败的那条约束，所以两个 issue 收敛于同一个定义，而不是两个恰好相关的修复。
+
+第二条更窄的规则用于识别整行粘贴的环境变量：匹配 `^[A-Z][A-Z0-9_]*=[^=]` 或首尾成对引号的输入会被拒绝。把前缀限定为全大写可以让真实 Key 与之绝缘——`sk-` 这类形态会在连字符处中断标识符匹配——而要求分隔符之后必须是非 `=` 字符，则让 base64 的 padding 也与之绝缘。它报出的是与非法字符相同的那条格式失败，而不是自己的一句：读到它的人下一步动作完全一样，因此单列一句只会点出一个原因，却不改变该怎么做。
+
+### 不变量属于每一层，启发式属于人所在的那一层
+
+字符集规则是不变量。非 ASCII 字符对任何 provider 都**不可能**在 header value 中传输，因此在浏览器、在各个 resolver、在每一次凭据读取上执行它，是结构上的一致而非约定上的一致。
+
+形状规则是对人如何粘贴的猜测，因此**只在浏览器中运行**。`llm-pi-ai` 前面挂着 OpenAI、Anthropic 以及任意手工声明的网关，本仓库并不掌握它们的 Key 格式；若这条规则运行在 resolver 中，一个签发形如 `TENANT1=abc` 的网关会让用户被彻底锁死、无路可走——设置页拒绝它，手写的 `.env` 在读取时同样被拒。把启发式限制在粘贴动作发生的那一层，环境变量便始终是那条出路。
+
+### 「没有 Key」是一种配置状态，不是缺失
+
+在这里，「没有 API Key」意味着三件完全不同的事，其中只有一件是错误。规则作用于**已提供**的值；至于究竟有没有提供，由各个调用方自行判断。
+
+**未指定。** 既不写 `apiKey` 也不写 `apiKeyEnv` 的 profile，是由 harness 所持有的 Key 之外的东西来鉴权的。[provider.ts](../../../../packages/llm/llm-pi-ai/src/provider.ts) 中的 `routeAuth` 保留内置 catalog provider 自身的鉴权，正是为了让 provider 原生的 ambient 发现得以存活；而该 catalog 附带的 `openai-codex` 通过 OAuth 鉴权，并会直接拒绝一个显式的 Key。`namesCredential` 承载着这一区分。在 `llm-deepseek` 中，缺省的 `apiKey` 同样会回落到 `apiKeyEnv`。未指定的情形永不参与校验。
+
+**Web UI 中留空的输入框。** 即便某个 provider 的 Key 已经存好，该输入框也是空着打开的——`keyStored` 的文案写的是「已配置——输入新值以替换」——所以留空意味着*保持已存储的值*。`ProviderEditor` 在草稿为空时完全跳过 `credentials.set`，这一点保持不变：留空绝不拦截提交，否则改一个 base URL 都得重新输一遍 Key。
+
+**已提供，但为空或纯空白。** 它意味着什么，取决于「缺失」在该界面上选中了什么，而两个适配器的差异是有依据的。在 `llm-pi-ai` 中它是错误，因为那里的缺失切换的是**鉴权方式**——转向内置 provider 的 ambient 发现或 OAuth——因此一个空 Key 究竟想选哪一种是真有歧义；它的措辞指明了合法替代路径而非单纯拒绝（*has an empty apiKey; omit it to use ambient authentication*）。在 `llm-deepseek` 中，缺失只是为同一把 Key 选择了另一个**来源** `apiKeyEnv`，因此空白字面量会像缺省一样经该回落解析。在浏览器中它始终是失败，两张卡片皆然：字段是人刚刚敲过字的地方，静默丢弃他敲进去的内容永远不是正确答案。
+
+因此 `normalizeApiKey` 接受 `string`，而绝非 `string | undefined`。
+
+### 规则住在哪里
+
+`normalizeApiKey` 是 `dsh-llm` seam 的一个模块，与已经承担共享 header 事务的 [attribution.ts](../../../../packages/llm/llm/src/attribution.ts) 并列。两个适配器都依赖该 seam 且都需要这条规则，因此它拥有两个当前消费者而非一个预设消费者。它返回 trim 后的值，或一个原因（`empty`、`illegalCharacters`）。
+
+两个适配器同样都需要那句完全相同的「拒绝一个已存储凭据」的诊断，差别仅在包名前缀。`LlmError` 声明在 seam 的 `index.ts` 中，因此 `assertUsableApiKey(raw, pkg, ref)` 就住在它旁边，两个适配器都不再各留一份。断言模块本身保持零依赖：把 `LlmError` 引入 `api-key.ts` 会与 `index.ts` 对它的再导出成环。
+
+客户端无法引入其中任何一个：client 包只 reference client 包，因此 `packages/client/ui-models` 在自己的 `apiKey.ts` 中镜像这个断言并持有本地化文案，正如 `validateDeepSeekModels` 镜像 host 侧的 `catalogModel` schema。两侧在注释中互相指名。
+
+### 各个界面各做什么
+
+| 界面 | 行为 |
+|---|---|
+| `dsh-llm` | 拥有 `normalizeApiKey`、`assertUsableApiKey` 与 `INVALID_CREDENTIAL_CODE`，后者刻意不进 `DEFAULT_RETRYABLE_CODES`。 |
+| `llm-deepseek` `resolveAdapterOptions` | 拒绝标头无法承载的字面量 `apiKey`，与其他超出 schema 的边界检查并排；使用 trim 后的值。缺省或空白的 `apiKey` 回落到 `apiKeyEnv`。 |
+| `llm-deepseek` `resolveApiKey` | 归一化凭据 seam 或环境返回的值，以 `INVALID_CREDENTIAL` 拒绝，消息指明模型设置页，绝不回显 Key。 |
+| `llm-pi-ai` `resolveProfiles` | 施加这条共享规则，保留其「omit it to use ambient authentication」的措辞，并把 trim 后的值写进解析后的 profile。 |
+| `llm-pi-ai` `resolveApiKey` | 归一化凭据与环境路径。不指定任何凭据的 profile 仍返回 `undefined`，ambient 与 OAuth 路由不受影响。 |
+| `llm-pi-ai` `discoverModels` | 在构造 header 之前归一化，使非法 Key 成为凭据故障而非端点不可达。不带 Key 的探测保持未鉴权。 |
+| `ui-models` | 镜像字符集规则，加入形状启发式，在探测与 `credentials.set` 之前 trim `keyDraft`，并修正 `stringAt` 的空值判断。留空的输入框仍是可以提交的空操作；只含空白的输入框则是字段级失败。提交**与端点探测**同时受拦截，因此被拒绝的密钥不会白花一次往返去换取字段上已经写明的答案；失败呈现在字段上，与既有的 `modelFailure` 模式一致。 |
+
+`ProviderEditor` 同时服务 DeepSeek 与 pi-ai 两种布局，因此一处客户端改动覆盖两个 provider。`CustomProviderCard` 为手工声明的路由承载同一套判定。
+
+`credentials-local` 刻意不动。它存储各类凭据，而可打印 ASCII 是 HTTP header 的约束而非凭据存储的约束；它既有的、拒绝任何 dotenv 样式都无法表示的值的行为保持原样。
+
+## Alternatives considered
+
+**在 `apiKey` schema 字段上加 `.pattern()`。** vendor 中的 schemastery 支持它，且该 pattern 会随命名空间 schema 一同序列化到浏览器——一条规则，投递而非镜像。它落败于 pattern 无法先行 trim：那样 `cordis.yml` 会拒绝带首尾空白的 Key 而 `.env` 却容忍，resolver 与 schema 会对同一个字符串给出分歧。在 `resolveAdapterOptions` 中校验可以让每一层都是 trim-then-validate，而该函数本就是本包重新裁定 schema 无法表达的边界之处。
+
+**由 client 与 host 共享一个校验模块。** 被 source plane 布局否决：client 包只 reference client 包外加 `vendor/cordis` 与 `support/invariants`，把它放宽到够得着 host 包会撞上这一分割本就要隔开的两份 `Context` 合并。在两侧各镜像一行断言并各配一份测试，是此处的既定形态。
+
+**在 `llm-deepseek` 与 `llm-pi-ai` 中各留一个抛错 helper。** 最初的计划正是各留一份，差别仅在消息中的包名前缀，并配一个重复检测豁免来放行这一对。在实现之前即被否决：`LlmError` 声明在 seam 中，因此 seam 完全可以自己拥有这句诊断，而那里的一个豁免恰恰会掩盖它本要遮掩的重复。
+
+**在适配器的 `catch` 中嗅探 `TypeError`。** 这只是事后归类 ByteString 失败，header 构造本身仍无防护。它依赖 Node 错误消息的措辞，因而会随运行时版本静默失效；它也帮不到 `llm-pi-ai`——后者的请求 header 构造在 pi-ai SDK 内部。在交出 Key 之前就拒绝，则对两个适配器与探测路径同时有效。
+
+**在 `credentials-local.set` 中执行。** 它能一次性拦住所有写入方，包括手工编辑的文件。它落败于该 provider 存储各种类型的凭据，而一条源自 HTTP header 编码的规则并不属于它。
+
+**让形状启发式也在 resolver 中运行。** 更对称，且能拦住直接写进 `.env` 的整行环境变量。因上文所述的锁死风险而否决：resolver 中的一次误判会让用户无路可走，浏览器中的一次误判则仍留有环境变量这条路。
+
+**在保存时探测 provider 以证明 Key 可用。** 它能关掉来源真正开篇抱怨的那件事——保存报成功、第一轮才失败。因超出范围而否决，且在当时的代码上无法建成：对 pi-ai 恰好自带 catalog 的那些 provider，`discoverModels` 会在任何网络调用之前短路到内置 catalog，因而对 Key 什么都验证不了；而 DeepSeek 卡片根本没有探测。验证器的价值在于分清「Key 被拒」与「无法连通」，而这正是本次改动让其变得可靠的区分；先建验证器只会得到一个分不清自身结果的验证器。同类产品也不在保存时验证，因此保存时的阻断式网络调用会是一个意外行为，而非一处缺失。
+
+## Consequences
+
+格式错误的 Key 在持有它的那个字段上就被拒绝；格式错误的已存储 Key 以 `INVALID_CREDENTIAL` 失败，消息指明修复位置且不含 Key 的任何片段。由于该 code 位于 `DEFAULT_RETRYABLE_CODES` 之外，一个确定性的凭据故障不再被当作瞬时传输抖动重试三次。`llm-pi-ai` 的探测把非法 Key 报为凭据故障，而非端点不可达。
+
+形状启发式可能拒绝一个真实的 Key。最初的写法匹配任意「全大写标识符接 `=`」，评审指出其覆盖面比预期更宽：一个以 padding 结尾的全大写 base64 Key（`ABCD==`）会命中它并不像的赋值形态。要求分隔符之后必须是非 `=` 字符即可排除 padding——base64 的 padding 只出现在末尾。剩下的形态（大写名称、一个 `=`、然后是值）是已知 provider 不会签发的，且该规则只在浏览器中运行，因此仍撞上它的用户可通过环境变量设置该凭据。残留代价是对一个尚无人报告过的 Key 给出一次令人困惑的拒绝。
+
+限定为可打印 ASCII 比传输本身的要求更严：header value 是可以承载 `\x80`–`\xFF` 的。放行 latin-1 会让 `é` 通过并换回一个语焉不详的 401，而不是一次本地的、有解释的拒绝，因此从严是刻意的。若某个 provider 签发 latin-1 的 Key，这条规则需要放宽。
+
+字符集断言存在两份，每个 source plane 一份。布局禁止共享它；两侧各自带测试并在注释中指名其孪生体。
+
+早先版本已存下的 Key 会经 `resolveApiKey` 读取，因此一个非法的既存值将从解析时开始失败，而非到请求时才失败。诊断变好了，但对当前正持有这类值的人而言，失败点提前了。
+
+把这件事做错的最大代价，会是把「未指定」当成「非法」：一条施加到 `undefined` 上的规则会打断每一条依赖 ambient 发现或 OAuth 鉴权的路由，而一个会拦截提交的空输入框，则会让改动任何其他设置都必须重新输入 Key。这两点都由测试钉住，而不是仅仰赖谨慎。
+
+## Testing
+
+`packages/llm/llm/tests/api-key.spec.ts` 以整张输入表驱动 `normalizeApiKey` 与 `assertUsableApiKey`——空值、纯空白、带首尾空白、含中间空格、C0 控制字符、emoji、中文、全角、latin-1，以及可打印 ASCII 的边界字符——并钉住一次拒绝携带 `INVALID_CREDENTIAL` 且不含 Key 的任何部分。
+
+`packages/llm/llm-deepseek/tests/` 在 `adapter.spec.ts` 中覆盖字面量配置路径，在 `dynamic-config.spec.ts` 中经真实凭据 seam（而非 stub）端到端覆盖已存储凭据路径。`packages/llm/llm-pi-ai/tests/` 覆盖 `resolveProfiles`——包括 trim 后的值确实到达解析后的 profile，否则会被 `...rest` 展开丢弃——以及探测路径，包括不带 Key 的探测不会发出 `authorization` 标头。
+
+`packages/client/ui-models/tests/` 以同一张表加上形状用例钉住 `apiKeyFailure`，并驱动两张卡片：留空的输入框可提交且不写入凭据、只含空白的输入框在字段上失败、非法或被包裹的 Key 同时拦截提交与探测、带首尾空白的 Key 在 `credentials.set` 与探测之前被 trim，以及手工声明的路由可以完全不带 Key 创建。
+
+用户可见的终态则钉在它真正被组装的位置：`examples/headless-agent/tests/headless.snapshot.ts` 让 one-shot 应用在一个 HTTP 标头无法承载的已存密钥下运行，复用其 missing-credential 兄弟场景的同一套无密钥 composition，并记录该轮以 `INVALID_CREDENTIAL` 结束、消息可操作且既不含密钥也不含 `ByteString` 字样。包级测试无法证明这一点，而 web e2e 只覆盖了浏览器那一半。

.agents/notes/implemented/bug-fix/2026-08-06-onboarding-step-owned-takeover-chrome.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/2026-08-06-onboarding-step-owned-takeover-chrome.md
+2026-08-06-onboarding-step-owned-takeover-chrome.md: 4b3bbbc03c4494359297ae6e54bcc9a74c387e80
+2026-08-06-onboarding-step-owned-takeover-chrome.zh.md: 548285d285939325a26df3b8d70c43fe952813a9

.agents/notes/implemented/bug-fix/2026-08-06-onboarding-step-owned-takeover-chrome.md

@@ -0,0 +1,35 @@
+# Agent Note: onboarding takeover chrome moves into the step
+
+Status: implemented
+
+English | [中文](2026-08-06-onboarding-step-owned-takeover-chrome.zh.md)
+
+## Problem
+
+The settings shell mounted the onboarding takeover chrome — a body-portaled overlay with an opaque `--dsw-alias-bg-layer-1` stage, a blur mask, and `#root` set inert — the moment a `settings.onboarding` step was registered and not yet locally completed. Every step decides whether it actually needs to show by loading a private fact first (WelcomeNotice: the acknowledgement bit through its settings join; DeepSeekOnboardingDialog: credential readiness through the Models join) and renders `null` while that fact is in flight. Rendering `null` could not suppress the chrome, because the opaque stage was painted by the shell around the slot outlet, not by the step.
+
+On every reload while the hero (blank or no session) was current, the sessions list turning `ready` therefore popped a full-screen opaque layer — white in the light palette — and blocked all interaction for exactly one credential/settings RPC round-trip, after which the already-configured steps self-completed and the layer vanished. Users saw the app flash white each refresh the moment the workspace/session lists landed.
+
+## Decision
+
+The takeover chrome belongs to the step, not the shell. A new zero-cordis primitive, `OnboardingSurface` (ui-primitives), renders the body-portaled overlay/mask/stage — CSS class names and geometry moved verbatim from `SettingsRoot.module.css` — and holds `#root` inert for exactly its own mount lifetime. Both step components wrap only their **visible** branch in it; their existing `null` branches now paint and block nothing by construction, because the chrome is part of the same render decision.
+
+`SettingsRoot` keeps the coordinator exactly as it was (ordered ledger projection, one mounted step, local completed set, `stepId`/`complete`/`openSection` currency) but renders the elected step bare — no portal, no stage, no inert effect. The `settings.onboarding` slot contract now states that registrants own the surface wrap and must render `null` while their private facts are undecided.
+
+## Alternatives considered
+
+**Register steps conditionally (ledger as the has-content signal).** Register the entry only after the private join resolves to "needs intervention". Architecturally clean (publish at the commit point) but a larger change: the join load must move from the dialogs into each plugin's apply, and registration/disposal becomes reactive plumbing in two packages. Rejected as oversized for the defect.
+
+**Convert `settings.onboarding` to a chain with an externalized completed-set store.** The composer-takeover pattern; prototyped and reverted. Selectors can only judge owner props, so the private readiness facts still had to be resolved inside the components — the chain bought routing generality the two current steps do not need, at the cost of a contract change across three packages.
+
+**Detect empty slot output at the render site.** `renderSlot` returns an outlet element unconditionally, so the owner cannot branch on a step's `null`; probing rendered DOM emptiness needs a commit-then-retract dance whose dynamic transitions lose the pre-paint guarantee.
+
+## Consequences
+
+While a step is mounted but undecided, the application stays visible and interactive: `#root` is no longer inert during the decision window (previously it was inert behind an opaque layer). For a genuinely unconfigured user the takeover now appears one join round-trip later than before — but with its content already present, instead of an empty stage that fills in.
+
+A future step that registers without wrapping its visible content in `OnboardingSurface` renders bare over the app with no mask; the slot contract JSDoc names the wrap as the registrant's obligation.
+
+## Testing
+
+`packages/client/ui-primitives/tests/onboarding-surface.spec.tsx` pins the primitive: body portal around the content, mask/stage class presence, `#root` inert held for exactly the mount lifetime, and the no-`#root` composition. `packages/client/ui-settings/tests/settings-root.spec.tsx` pins the inverted shell contract: no takeover chrome and no inert while a mounted step renders nothing. `apps/web/tests/onboarding-deepseek-config.e2e.ts` gains the defect's assembled regression pin: a configured world reloads while every `settings.describe` response is held open at the browser's network boundary — widening the steps' deciding window from loopback-invisible to hundreds of milliseconds, which is what keeps the assertions non-vacuous — and an 8 ms in-page sampler proves the takeover chrome never mounts and `#root` never turns inert. The file's existing scenarios and the step specs (`ui-settings-general`, `ui-models`) pass unchanged — the mask selector and geometry pins survive because the stylesheet moved verbatim.

.agents/notes/implemented/bug-fix/2026-08-06-onboarding-step-owned-takeover-chrome.zh.md

@@ -0,0 +1,35 @@
+# Agent Note：首次使用引导的接管界面框架移入步骤自身
+
+状态：已实现
+
+[English](2026-08-06-onboarding-step-owned-takeover-chrome.md) | 中文
+
+## 问题
+
+设置外壳在 `settings.onboarding` 有已注册且本地未完成的步骤时，就立即挂出首次使用引导的接管界面框架——portal 到 body 的浮层，带不透明的 `--dsw-alias-bg-layer-1` 展示层、模糊遮罩，并把 `#root` 置为 `inert`。而每个步骤都要先加载私有事实才能判定自己是否需要出场（WelcomeNotice：经其设置 join 读取确认位；DeepSeekOnboardingDialog：经 Models join 读取凭据就绪状态），判定期间渲染 `null`。渲染 `null` 无法抑制界面框架，因为不透明展示层是外壳画在 slot outlet 外面的，不属于步骤。
+
+于是每次在 hero（空白或无会话）状态下刷新页面，会话列表一变 `ready` 就弹出整屏不透明层——亮色主题下是白色——并阻断全部交互，时长恰好等于一次凭据/设置 RPC 往返；之后已配置好的步骤自我完成，图层消失。用户看到的就是每次刷新在 workspace/会话列表落地的瞬间闪一下白屏。
+
+## 决定
+
+接管界面框架属于步骤，不属于外壳。新增零 cordis 原语 `OnboardingSurface`（ui-primitives）：渲染 portal 到 body 的浮层／遮罩／展示层——CSS 类名与几何从 `SettingsRoot.module.css` 逐字迁移——并在自身挂载生命周期内保持 `#root` 为 `inert`。两个步骤组件只把各自的**可见**分支包进该原语；既有的 `null` 分支由此在构造上不绘制、不阻塞任何内容，因为界面框架已是同一次渲染决策的一部分。
+
+`SettingsRoot` 的协调器原样保留（有序账本投影、每次挂载一个步骤、本地完成集合、`stepId`／`complete`／`openSection` currency），但对当选步骤裸渲染——不再有 portal、展示层和 inert 效果。`settings.onboarding` 的 slot 契约现在写明：注册方持有外层包裹，且在私有事实未决时必须渲染 `null`。
+
+## 曾考虑的替代方案
+
+**条件注册（账本即有内容信号）。** 私有 join 解析出「需要介入」后才注册条目。架构上干净（在 commit point 发布），但改动更大：join 的加载必须从对话框上移到各插件的 apply，注册／销毁在两个包里都变成响应式接线。对本缺陷而言过重，否决。
+
+**把 `settings.onboarding` 改成 chain 并把完成集合外置为 store。** composer takeover 的版型；做过原型后回退。selector 只能判定 owner props，私有就绪事实仍然只能在组件内部解析——chain 买来的是当前两个步骤并不需要的路由通用性，代价却是跨三个包的契约变更。
+
+**在渲染点探测 slot 输出为空。** `renderSlot` 无条件返回 outlet 元素，owner 无法据步骤的 `null` 分支；探测已渲染 DOM 是否为空需要先提交再撤回的手法，其动态翻转会失去 paint 前的保证。
+
+## 后果
+
+步骤已挂载但尚未判定期间，应用保持可见且可交互：判定窗口内 `#root` 不再是 `inert`（此前是在不透明图层背后被置灰）。对真正未配置的用户，接管层比从前晚一个 join 往返出现——但一出现就带着内容，而不是先露出空白展示层再填充。
+
+未来若有步骤注册后不把可见内容包进 `OnboardingSurface`，会无遮罩地裸渲染在应用之上；slot 契约的 JSDoc 已把包裹写为注册方的义务。
+
+## 测试
+
+`packages/client/ui-primitives/tests/onboarding-surface.spec.tsx` 钉住原语行为：内容外的 body portal、遮罩／展示层类名存在、`#root` 的 `inert` 恰好持续挂载生命周期，以及无 `#root` 的组合。`packages/client/ui-settings/tests/settings-root.spec.tsx` 钉住反转后的外壳契约：已挂载步骤什么都不渲染时，无接管界面框架、无 inert。`apps/web/tests/onboarding-deepseek-config.e2e.ts` 新增本缺陷的整装回归钉：已配置世界刷新页面，同时在浏览器网络边界扣住所有 `settings.describe` 响应——把步骤的判定窗口从 loopback 下不可见拉宽到数百毫秒，这正是断言保持非空洞的关键——页内 8ms 采样器证明接管界面框架从未挂载、`#root` 从未变为 inert。该文件的既有场景与步骤 spec（`ui-settings-general`、`ui-models`）原样通过——样式表逐字迁移，遮罩选择器与几何钉子得以幸存。

.agents/notes/implemented/bug-fix/2026-08-06-provider-credential-lifecycle.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/2026-08-06-provider-credential-lifecycle.md
+2026-08-06-provider-credential-lifecycle.md: ce45207e7ac7224f44e34945e36ba85db0971f09
+2026-08-06-provider-credential-lifecycle.zh.md: c476417517b8ed72036344a13720a8ba378775e6

.agents/notes/implemented/bug-fix/2026-08-06-provider-credential-lifecycle.md

@@ -0,0 +1,27 @@
+# Agent Note: Recoverable provider credential lifecycle
+
+Status: implemented
+
+English | [中文](2026-08-06-provider-credential-lifecycle.zh.md)
+
+## Problem
+
+The Models editor spans independent settings and credential RPC domains. It previously committed provider settings before storing the API key but kept the revision and original subtree from when the card opened. If the credential write failed, retry replayed the already-committed settings mutation with a stale revision and produced a conflict, leaving the user unable to complete the second stage from the same card. A blank pi-ai key also wrote the derived `apiKeyEnv` without a credential, which prevented pi-ai from using provider-native discovery. At deletion, the inverse leak remained: the profile disappeared but its page-stored key stayed in `.env` and silently became active when the provider was added again. Generic row actions and confirmation copy did not identify which provider would be changed.
+
+## Decision
+
+Provider save remains a two-stage settings-then-credentials operation over the existing wire domains, but the card treats the successful settings response as a commit checkpoint. It replaces its comparison subtree and expected revision with the returned redacted descriptor before attempting `credentials.set`; if that second stage fails, the draft key and card stay visible, and retry produces no settings ops and repeats only the credential write. Genuine concurrent changes before the first settings commit still fail with `settings-conflict`. Typed keys are trimmed at the UI and direct DeepSeek resolver boundaries, and pi-ai records a derived reference only when the normalized key is non-empty; saving a blank key materializes an empty, reference-free profile for provider-native discovery.
+
+Deletion removes a credential only when the joined row identifies the exact `<ROUTE>_API_KEY` reference derived by this page and reports it configured and writable. It unsets that credential before the user-layer profile so a settings-stage failure leaves the row and its frozen target visible for retry; both unsets are idempotent. Custom references, environment credentials, missing credentials, and targets the join cannot identify are retained. The row's accessible Edit/Delete names and the destructive dialog title, description, and final action all use the same stable `Display Name (route-id)` identity, collapsing to the route id when both strings match. The dialog states whether the stored key will be removed and owns operation failures instead of replacing the whole page with a load-error banner. Rows expose API-key state only from the value-free join: a confirmed literal or referenced credential is a green solid dot, a confirmed missing named reference is a red solid dot, and reference-free provider-native authentication or unavailable credential enrichment has no dot. Each dot has accessible copy and a tooltip, while successful Apply uses the same provider identity in a local status message and never echoes secret material.
+
+## Alternatives considered
+
+**Add a cross-domain transaction RPC.** Settings and credentials have separate owning services and durable stores; introducing a new host transaction would broaden the public wire and still require compensation for provider-specific persistence failures. The UI checkpoint makes the current ordered stages recoverable without adding a fourth configuration contract.
+
+**Delete every credential reference named by a removed profile.** A custom reference can be shared, externally managed, or intentionally survive profile churn. Exact equality with this page's derived target plus configured+writable state is the narrow evidence available to the page; anything weaker risks deleting a credential it does not own.
+
+**Remove settings first and compensate by recreating the profile.** The browser holds only a redacted subtree and cannot faithfully reconstruct stored literal secrets or concurrent edits. Credential-first deletion leaves the authoritative profile visible on partial failure and makes retry safe without synthesizing configuration.
+
+## Consequences
+
+The Models page can recover from either second-stage failure without reload, secret disclosure, or a false concurrency conflict, and blank-key pi-ai profiles preserve Bedrock, Vertex, and other provider-native authentication. Confirmed status is visible without turning route liveness, native authentication, or a failed credential lookup into a false error, and a successful replacement remains observable even when the row stays green. Deleting a page-managed provider no longer leaves a reusable local key, while ambiguous credentials deliberately remain for manual management. Save and delete are still not atomic across durable stores: a process crash can stop between stages, but their order and idempotence leave an observable, retryable state. Component tests pin partial-success retries, empty-key native auth, normalized literals, status visibility, target identity, cleanup ownership, and credential/settings rejection ordering; the keyless browser scenario pins bilingual accessible copy and verifies that confirmed deletion removes both `settings.yaml` profile and `.env` credential. This decision refines the Models apply semantics recorded in the [web configuration plane note](../architecture/2026-07-30-web-config-plane.md).

.agents/notes/implemented/bug-fix/2026-08-06-provider-credential-lifecycle.zh.md

@@ -0,0 +1,27 @@
+# Agent Note: 可恢复的提供方凭据生命周期
+
+Status: implemented
+
+[English](2026-08-06-provider-credential-lifecycle.md) | 中文
+
+## 问题
+
+Models 编辑器横跨互相独立的 settings 与凭据 RPC 领域。之前它先提交提供方 settings，再存储 API 密钥，却一直保留卡片打开时的 revision 和原始子树。如果凭据写入失败，重试会用陈旧 revision 重放已提交的 settings 变更，并产生冲突，导致用户无法从同一张卡片完成第二个阶段。空的 pi-ai 密钥还会写入派生的 `apiKeyEnv`，却不写入凭据，从而阻止 pi-ai 使用提供方原生凭据发现。删除时则存在相反的残留问题：profile 消失了，页面存储的密钥却保留在 `.env` 中，并在重新添加提供方时静默地恢复作用。笼统的行操作与确认文案也没有标明要更改哪个提供方。
+
+## 决策
+
+提供方保存仍在现有 wire 领域上按先 settings、后凭据的两阶段顺序执行，但卡片会把成功的 settings 响应视为提交检查点。它会在尝试 `credentials.set` 之前，用返回的脱敏 descriptor 替换比较基准子树与预期 revision；如果第二阶段失败，草稿密钥与卡片会继续显示，重试不会产生 settings op，只会再次写入凭据。首次 settings 提交之前发生的真实并发变更仍会以 `settings-conflict` 失败。UI 与 DeepSeek 直连 resolver 边界均会去除所输密钥的首尾空白，且只有标准化密钥非空时，pi-ai 才会记录派生引用；留空密钥会具化一个空的、不带引用的 profile，以便使用提供方原生凭据发现。
+
+只有当联接所得的行识别出该页面派生的精确 `<ROUTE>_API_KEY` 引用，并将其报告为已配置且可写时，删除操作才会清除该凭据。它会先取消设置该凭据，再取消设置用户层 profile；如果 settings 阶段失败，该行及其已冻结的目标仍可见，便于重试。两项 unset 都具备幂等性。自定义引用、环境凭据、缺失的凭据，以及联接无法识别目标的凭据均会保留。行的无障碍 Edit/Delete 名称以及破坏性对话框的标题、说明和最终操作都使用同一个稳定的 `Display Name (route-id)` 标识；当两个字符串相同时，标识会简化为路由 id。对话框会说明是否一并删除已存密钥，并在自身内显示操作失败，而不是用加载错误横幅替换整个页面。行只根据不含值的联接结果展示 API 密钥状态：确认已配置的字面密钥或引用凭据显示为绿色实心点，确认缺失的具名引用显示为红色实心点，无引用的提供方原生认证或无法取得凭据补充信息时则不显示状态点。每个状态点都有无障碍文案和工具提示；「应用」成功后的本地状态消息会使用同一个提供方标识，且绝不回显任何机密内容。
+
+## 曾考虑的替代方案
+
+**添加跨领域事务 RPC。**settings 与凭据分属不同的主管服务与持久存储；引入新的 Host 事务会扩大公开 wire 面，而且仍需要补偿提供方特定的持久化失败。UI 检查点让当前的有序阶段变得可恢复，无需添加第四项配置契约。
+
+**删除被移除 profile 所指定的每一个凭据引用。**自定义引用可能被共享、由外部管理，或有意在 profile 反复增删时存留。与该页面派生目标精确相等，再加上已配置且可写的状态，是页面所能获得的最小范围证据；比这更弱的判定都有可能删除不属于它的凭据。
+
+**先删除 settings，再重建 profile 以作补偿。**浏览器只持有脱敏后的子树，无法忠实重建已存的字面机密或并发编辑。先删除凭据可以让权威 profile 在部分失败时仍然可见，并且无需合成配置就能安全重试。
+
+## 后果
+
+Models 页可以从任一第二阶段失败中恢复，无需重新加载，也不会泄露机密或产生虚假的并发冲突；空密钥的 pi-ai profile 会保留 Bedrock、Vertex 与其他提供方原生认证。已确认的状态清晰可见，同时不会把路由存活状态、原生认证或凭据查询失败误报为错误；即使该行继续显示绿色，密钥替换成功也仍然可观察。删除由页面管理的提供方不再遗留可重用的本地密钥，而存在歧义的凭据会有意保留，交由手动管理。保存与删除在跨持久存储时仍非原子操作：进程可能在两个阶段之间崩溃，但它们的顺序与幂等性会留下可观察、可重试的状态。组件测试固定了部分成功后的重试、空密钥原生认证、标准化字面值、状态可见性、目标标识、清理所有权，以及凭据／settings 拒绝顺序；无密钥的浏览器场景固定了双语无障碍文案，并验证确认删除会同时清除 `settings.yaml` profile 与 `.env` 凭据。此决策细化了 [web 配置平面 note](../architecture/2026-07-30-web-config-plane.md) 中记录的 Models 应用语义。

.agents/notes/implemented/bug-fix/2026-08-06-reader-scroll-attribution-observed-top-ledger.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/2026-08-06-reader-scroll-attribution-observed-top-ledger.md
+2026-08-06-reader-scroll-attribution-observed-top-ledger.md: 2dcd0178e4216d4d3e2edcd3bfcc05b63606c333
+2026-08-06-reader-scroll-attribution-observed-top-ledger.zh.md: ffb03fdb68c9cb040e311077dd698b310a033fdd

.agents/notes/implemented/bug-fix/2026-08-06-reader-scroll-attribution-observed-top-ledger.md

@@ -0,0 +1,37 @@
+# Agent Note: Reader scroll attribution through the observed-top ledger
+
+Status: implemented
+
+English | [中文](2026-08-06-reader-scroll-attribution-observed-top-ledger.zh.md)
+
+## Problem
+
+ChatView's bottom-follow recognized only wheel/trackpad gestures as reader input: while pinned to the floor, a scroll event without matching wheel movement was treated as programmatic and snapped back. Touch panning, native-scrollbar dragging, and keyboard paging therefore could not leave the bottom of a streaming transcript — on a phone the tail was effectively locked. That wheel-only provenance was a deliberate deferral in the [sticky-composer note](2026-07-29-sticky-composer-conversation-scroll.md), which rejected a general input state machine "for this narrow fix" and left every other scroll source outside the model.
+
+## Decision
+
+Reader input is no longer identified by device. ChatView keeps an observed-top ledger (`observedTopRef`): the last `scrollTop` either delivered on the main thread or written by the component, recorded synchronously at every programmatic write site — bottom follow, open restore, prepend anchoring, resize follow, and scroll delivery itself. When a scroll event arrives, a position that deviates from `min(ledger, floor)` by more than half a pixel is reader input; a position on the ledger (a delayed programmatic delivery) or exactly on the shrunken floor (a browser clamp after content shrank) preserves the current ownership state. Ownership then changes only through reader input under the existing threshold rule: within `FOLLOW_THRESHOLD` of the floor re-pins, beyond it releases follow and shows Back to bottom. The wheel listener and its epoch bookkeeping are deleted; the component listens to `scroll` alone, so wheel, touch, scrollbar, keyboard, and any future input source are covered by one rule.
+
+## Contract change: coalesced shrink-plus-regrow clamps
+
+A shrink clamp whose layout regrows within the same rendering update before the clamp's scroll event is delivered is geometrically indistinguishable from reader input, so it now reads as the reader and releases follow (Back to bottom recovers). Realistic React-commit-driven shrink and regrow is still absorbed: the layout-effect follow re-pins and re-records the ledger per commit, and a shrink-only clamp lands exactly on `min(ledger, floor)`. Only a non-React reflow that shrinks and regrows inside one update mis-attributes. The previous wheel model kept following in that raced case; the unit contract was rewritten to the absorbed-shrink-only guarantee in the same change.
+
+## Testing
+
+Unit specs in `packages/client/ui-conversation/tests/chat-view.spec.tsx` pin the ledger contract directly: a `readerScroll` helper delivers a position the component never wrote, programmatic deliveries land on the ledger, and the stream-finalization shrink clamp keeps following. Two scenarios in `apps/web/tests/chat-scroll-contract.e2e.ts` extend the [browser e2e lane](../testing/2026-07-24-web-gui-browser-e2e-lane.md): keyboard paging over a settled transcript and a touch-style momentum fling against paced streaming, both red under the wheel-only implementation and green under the ledger.
+
+The lane's Chromium cannot synthesize any non-wheel device scrolling, which bounds what the e2e can drive for real: `Input.synthesizeScrollGesture` with a touch source and hand-rolled `Input.dispatchTouchEvent` sequences deliver DOM events but never move a scroller (headless and headed-under-Xvfb alike); the `default` gesture source synthesizes wheel events; and compositor scrollbars ignore synthetic mouse input entirely, with a gutter visible only when `--hide-scrollbars` is removed. Keyboard is the one working non-wheel primitive, so it carries the real-input-pipeline proof, and the fling scenario replays touch's signature — per-frame decaying displacements the component never authored — through the scrollport directly.
+
+## Alternatives considered
+
+**Keep the wheel-only model.** Rejected: it is the defect. Touch, scrollbar, and keyboard readers cannot take ownership away from a streaming tail, and each newly supported device would need its own carve-out.
+
+**Enumerate input devices.** Adding `touchstart`/`pointerdown`/`keydown` listeners beside the wheel epoch was the obvious extension. Rejected: native-scrollbar dragging exposes no input event to latch before its scrolls arrive, device lists rot as browsers add sources, and every listener would need its own compositor-delivery grace window — the input state machine the sticky-composer note already declined to build.
+
+**Absorb the coalesced shrink-plus-regrow clamp with heuristics.** Floor-mismatch grace windows or deferred rAF re-checks could keep the raced clamp from reading as the reader. Rejected: streaming rewrites the floor at chunk pace (24 ms) against ~16 ms frames, so any grace window either swallows genuine touch input during streaming — reopening the bug this change fixes — or is too short to cover the race it targets. The mis-attribution is accepted and recoverable instead.
+
+**Drive real touch and scrollbar devices in e2e.** Rejected by the environment, not by preference: every synthesis path (CDP touch gestures, touch event sequences, synthetic mouse on classic scrollbars, headed under Xvfb) was probed and cannot scroll; the details live in Testing above.
+
+## Consequences
+
+Every reader input owns bottom-follow uniformly, with less code: the wheel listener, its epoch counter, and the pre-input baseline bookkeeping are gone, and attribution rides state the component already maintained. The sticky-composer note's layout, wheel chaining, and prepend-anchoring decisions are untouched and remain authoritative; its narrow provenance rule is superseded by this note. The cost is the contract change above — a coalesced non-React shrink-plus-regrow clamp now pauses follow until the reader returns to the floor or presses Back to bottom — traded for touch, scrollbar, and keyboard correctness during streaming. The e2e lane gains non-wheel coverage only within what its browser can synthesize; if gesture synthesis starts working in a future Chromium, the fling emulation can be replaced by real touch strokes without changing the asserted contract.

.agents/notes/implemented/bug-fix/2026-08-06-reader-scroll-attribution-observed-top-ledger.zh.md

@@ -0,0 +1,37 @@
+# Agent Note: 经由 observed-top ledger 的读者滚动归因
+
+Status: implemented
+
+[English](2026-08-06-reader-scroll-attribution-observed-top-ledger.md) | 中文
+
+## 问题
+
+ChatView 的贴底跟随此前只把滚轮／触控板手势识别为读者输入：钉在底部（floor）期间，一个没有对应滚轮位移的滚动事件会被视为程序化滚动并被拉回底部。因此触控平移、拖动原生滚动条与键盘翻页都无法离开流式 transcript（文本记录）的底部，在手机上尾部实际上被锁死。这种仅认滚轮的输入来源判定是 [sticky-composer 笔记](2026-07-29-sticky-composer-conversation-scroll.md)中有意的暂缓：该笔记拒绝为「此次窄范围修复」建立通用输入状态机，把其余所有滚动来源都留在模型之外。
+
+## 决策
+
+读者输入不再依据设备来识别。ChatView 维护一份 observed-top ledger（`observedTopRef`）：即最近一次由主线程交付、或由组件自身写入的 `scrollTop`，并在每一个程序化写入点（贴底跟随、打开时恢复、前置锚定、尺寸变化跟随以及滚动交付本身）同步记录。滚动事件到达时，偏离 `min(ledger, floor)` 超过半像素的位置即为读者输入；落在 ledger 上的位置（迟到的程序化交付），或恰好落在收缩后底部上的位置（内容收缩后的浏览器钳制），则维持当前的所有权状态。此后所有权只经由读者输入、按既有阈值规则变化：位置距底部在 `FOLLOW_THRESHOLD` 以内则重新贴底，超出则释放跟随并显示「回到底部」。滚轮监听器及其 epoch 簿记已删除；组件只监听 `scroll`，因此滚轮、触控、滚动条、键盘以及未来任何输入来源都由同一条规则覆盖。
+
+## 契约变更：收缩与重新增长被合并的钳制
+
+如果一次收缩钳制的布局在同一次渲染更新内、赶在该钳制的滚动事件交付之前重新增长，那么这个事件在几何上与读者输入无法区分，因此它现在会被判读为读者并释放跟随（可经「回到底部」恢复）。现实中由 React 提交（commit）驱动的收缩与重新增长仍会被吸收：layout effect 中的跟随会在每次提交后重新贴底并重新记录 ledger，而仅收缩的钳制会恰好落在 `min(ledger, floor)` 上。只有在同一次更新内先收缩再重新增长的非 React 重排会被误归因。旧的滚轮模型在这种竞态情形下会保持跟随；单元测试契约已在同一变更中改写为只吸收纯收缩的保证。
+
+## 测试
+
+`packages/client/ui-conversation/tests/chat-view.spec.tsx` 中的单元测试直接钉住 ledger 契约：`readerScroll` 辅助函数交付一个组件从未写入过的位置，程序化交付落在 ledger 上，流收尾阶段的收缩钳制保持跟随。`apps/web/tests/chat-scroll-contract.e2e.ts` 中的两个场景扩展了[浏览器 e2e 车道](../testing/2026-07-24-web-gui-browser-e2e-lane.md)：在已停稳的 transcript 上做键盘翻页，以及对着按节奏推进的流式输出做一次触控式惯性快滑（momentum fling）；两者在仅认滚轮的实现下均为红、在 ledger 下均为绿。
+
+该车道的 Chromium 无法合成任何非滚轮的设备滚动，这限定了 e2e 能真实驱动的范围：触控来源的 `Input.synthesizeScrollGesture` 与手工构造的 `Input.dispatchTouchEvent` 序列都能交付 DOM 事件，却从不移动滚动容器（无头模式与 Xvfb 下的有头模式皆然）；`default` 手势来源合成的是滚轮事件；合成器滚动条则完全无视合成的鼠标输入，且只有移除 `--hide-scrollbars` 后才能看到滚动条槽。键盘是唯一可用的非滚轮原语，因此由它承担真实输入流水线的证明；快滑场景则把触控的特征（组件从未写入过的逐帧衰减位移）直接回放进滚动容器。
+
+## 曾考虑的替代方案
+
+**保留仅认滚轮的模型。** 否决：它本身就是缺陷所在。触控、滚动条与键盘读者无法从流式尾部夺走所有权，而每种新支持的设备都需要单独开一个特例。
+
+**逐一枚举输入设备。** 在滚轮 epoch 旁边加挂 `touchstart`/`pointerdown`/`keydown` 监听器是最显而易见的扩展。否决：拖动原生滚动条在其滚动事件到达之前不暴露任何可供锁存的输入事件；设备清单会随浏览器新增输入来源而逐渐腐化；而且每个监听器都需要自己的合成器交付宽限窗口——这正是 sticky-composer 笔记当初就拒绝构建的那个输入状态机。
+
+**用启发式吸收收缩与重新增长被合并的钳制。** 针对底部失配的宽限窗口，或推迟到 rAF 的复查，本可让这种竞态下的钳制不被判读为读者。否决：流式输出以分片节奏（24 ms）改写底部，而帧间隔约 16 ms，因此任何宽限窗口要么会在流式输出期间吞掉真实的触控输入（重新打开本次变更所修复的缺陷），要么短到盖不住它想针对的竞态。转而接受这一误归因，它是可恢复的。
+
+**在 e2e 中驱动真实的触控与滚动条设备。** 否决来自环境，而非偏好取舍：每条合成路径（CDP 触控手势、触控事件序列、经典滚动条上的合成鼠标、Xvfb 下的有头模式）都逐一试过，均无法滚动；细节见上文「测试」一节。
+
+## 后果
+
+每种读者输入现在都以同一方式拥有贴底跟随，而代码更少：滚轮监听器、它的 epoch 计数器以及输入前基线簿记均已移除，归因搭载在组件本就维护的状态之上。sticky-composer 笔记中的布局、滚轮链式处理与前置锚定决策原样保留，仍为权威；其窄范围的输入来源规则由本笔记取代。代价就是上文的契约变更：一次收缩与重新增长被合并的非 React 钳制现在会暂停跟随，直到读者回到底部或按下「回到底部」；以此换来流式输出期间触控、滚动条与键盘的正确性。e2e 车道获得的非滚轮覆盖仅限其浏览器能够合成的范围；若手势合成在未来某个 Chromium 版本中开始可用，可以在不改变所断言契约的前提下，把快滑模拟替换为真实的触控划动。

.agents/notes/implemented/bug-fix/2026-08-06-token-surface-unpriced-replace-compatibility.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/2026-08-06-token-surface-unpriced-replace-compatibility.md
+2026-08-06-token-surface-unpriced-replace-compatibility.md: 77b778fba695f560eb68af6da2f416e61ccff181
+2026-08-06-token-surface-unpriced-replace-compatibility.zh.md: dd4c5883c4c6d37809c2f8100f266655d8ba600c

.agents/notes/implemented/bug-fix/2026-08-06-token-surface-unpriced-replace-compatibility.md

@@ -0,0 +1,35 @@
+# Agent Note: unpriced surface replacements fold neutrally
+
+Status: implemented
+
+English | [中文](2026-08-06-token-surface-unpriced-replace-compatibility.zh.md)
+
+## Problem
+
+The `contextPressure` and `contextBreakdown` projections keep a running surface-token total plus at most one pending shadow-price claim, so their persisted checkpoints stay O(1) over a session's life. Current replace producers append a `compact/summary` or `compact/prune` metering event immediately before the replacement; its `shadowedTokenCount` prices the exact replaced range, and `foldSurfaceProjection` turns that into the signed delta.
+
+Sessions recorded before the shadow-price protocol log replacements with no adjacent metering event. The O(1) state cannot reconstruct the replaced range's price, and the fold treated every unpriced replacement as a contract violation and threw — so replaying such a session died at its first replacement (`token surface: replace at seq … has no adjacent shadow price`), leaving the session permanently unopenable.
+
+## Decision
+
+A replace that arrives with no armed claim folds price-neutrally: `foldSurfaceProjection` returns `deltaTokens: 0`, pricing the replaced range as if it had cost exactly what its replacement costs, and replay continues. A claim expired by an intervening event reaches the same neutral path, since the fold cannot distinguish it from a log that never metered.
+
+An armed claim naming a **different** range still throws. There the metering event was adjacent, so the producer wrote contradictory adjacent events — a live shadow-price contract violation, not historical data, and it must fail loud rather than let the total drift silently.
+
+Both projections share the one fold, so neither gains state fields nor bumps its `stateVersion`. `surface-fold.ts` and `ctx.tokenMeter.measure()` are unaffected: they hold the per-node priced surface and never needed the claim protocol.
+
+## Alternatives considered
+
+**Keep throwing.** Preserves the strict producer contract, but every pre-protocol session stays permanently unreplayable, and the projections exist to serve replay.
+
+**Persist the full priced surface in the projection state.** Could price any replaced range exactly, but grows the checkpoint by one node per model-visible message without bound — defeating the O(1) constraint the shadow-price protocol exists to preserve (see [the context-meter note](2026-08-05-context-meter-blind-to-compaction.md)).
+
+## Consequences
+
+An unpriced replacement holds the total still instead of shrinking it, so the compacted-away span stays counted: `contextBreakdown.messageTokens` retains the overcount, and `contextPressure.projectedTokens` overestimates occupancy only until the next usage sample re-anchors it, because that figure tracks movement since the sample rather than the absolute level. The error direction is safe — overestimating occupancy at worst invites an earlier compaction.
+
+The loud failure survives where it still means something: a range-mismatched adjacent claim is a current producer bug and still throws.
+
+## Testing
+
+`packages/llm/token-meter/tests/context-breakdown-projection.spec.ts` pins the neutral fold for the no-claim and expired-claim replacements, the throw for a mismatched claim, and the exact pricing for a matched one. `packages/llm/token-meter/tests/token-usage-projection.spec.ts` pins `contextPressure` holding still across an unpriced replacement.

.agents/notes/implemented/bug-fix/2026-08-06-token-surface-unpriced-replace-compatibility.zh.md

@@ -0,0 +1,35 @@
+# Agent Note: 未计价的表层替换以中性方式折叠
+
+Status: implemented
+
+[English](2026-08-06-token-surface-unpriced-replace-compatibility.md) | 中文
+
+## 问题
+
+`contextPressure` 与 `contextBreakdown` 两个投影只维护一份滚动累计的表层 token 总量，外加至多一条待结算的影子价格（shadow price）声明，因此其持久化检查点在会话整个生命周期内保持 O(1)。当前的替换生产方会紧贴在替换之前追加一条 `compact/summary` 或 `compact/prune` 计量事件；其 `shadowedTokenCount` 对被替换区间精确计价，`foldSurfaceProjection` 再把它换算成有符号增量。
+
+影子价格协议引入之前录制的会话，其日志中的替换没有相邻的计量事件。O(1) 状态无法重建被替换区间的价格，而折叠此前把每一次未计价替换都当作契约违规并抛出异常，于是回放这类会话会在第一处替换就中断（`token surface: replace at seq … has no adjacent shadow price`），会话从此永远无法打开。
+
+## 决策
+
+到达时没有已就位声明的替换以价格中性的方式折叠：`foldSurfaceProjection` 返回 `deltaTokens: 0`，相当于把被替换区间计价为恰好等于其替换内容的成本，回放随即继续。因中间插入的事件而过期的声明也走同一条中性路径，因为折叠无法把它与从未计量过的日志区分开。
+
+已就位但指向**另一个**区间的声明仍会抛出异常。此时计量事件确实相邻，说明生产方写入了互相矛盾的相邻事件：这是现行影子价格契约的违规，不是历史数据，必须响亮失败，而不能任由总量悄然漂移。
+
+两个投影共用同一个折叠，因此二者都不新增状态字段，也不提升 `stateVersion`。`surface-fold.ts` 与 `ctx.tokenMeter.measure()` 不受影响：它们持有逐节点的已计价表层，本来就不需要声明协议。
+
+## 备选方案
+
+**维持抛出异常。**保住了严格的生产方契约，但协议之前的每个会话都将永远无法回放，而投影本就是为服务回放而存在的。
+
+**在投影状态中持久化完整的已计价表层。**可以对任意被替换区间精确计价，但检查点会随每条模型可见消息各增加一个节点、无上限地增长，恰恰破坏了影子价格协议所要守住的 O(1) 约束（见[上下文仪表的 Agent Note](2026-08-05-context-meter-blind-to-compaction.md)）。
+
+## 影响
+
+未计价的替换让总量保持不动而不是缩小，因此被压缩（compaction）掉的区段仍被计入：`contextBreakdown.messageTokens` 保留这部分多计的量；`contextPressure.projectedTokens` 会高估占用率，但只持续到下一个用量样本重新锚定为止，因为该数字追踪的是自样本以来的增减，而非绝对水平。误差方向是安全的：高估占用率最坏不过是招致一次更早的压缩。
+
+响亮失败保留在它仍有意义的地方：区间不匹配的相邻声明是现行生产方的缺陷，仍会抛出异常。
+
+## 测试
+
+`packages/llm/token-meter/tests/context-breakdown-projection.spec.ts` 钉住了无声明与声明过期两种替换的中性折叠、声明区间不匹配时的抛出异常，以及声明匹配时的精确计价。`packages/llm/token-meter/tests/token-usage-projection.spec.ts` 钉住了 `contextPressure` 在一次未计价替换前后保持不动。

.agents/notes/implemented/feature/2026-06-15-code-mode.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/2026-06-15-code-mode.md
-2026-06-15-code-mode.md: b6a24ecd9700e32912b8112b59cbd8b6ab131eb5
-2026-06-15-code-mode.zh.md: a00a43ece1e581190de6096be8138df25a23f07f
+2026-06-15-code-mode.md: 99bbed3edab32512f88ece9694d6519a1f89c2dd
+2026-06-15-code-mode.zh.md: ca1bbe9ed3e412186763d1ed4fca9ed06669d4c3

.agents/notes/implemented/feature/2026-06-15-code-mode.md

@@ -6,7 +6,7 @@ English | [中文](2026-06-15-code-mode.zh.md)
 
 ## Problem
 
-In the registry's native presentation, the agent loop advertises every visible capability as a JSON-schema function definition. `ToolRegistry` contributes its schemas to the system-prompt assembly, the assembly's `tools` land on the wire (and in the logged request header), the model invokes one `tool-call` block per step, and the loop dispatches each call through `ctx.tools.execute()` **sequentially** (parallel tool execution is an explicit open TODO in `dsh-tools` and [docs/architecture.md](../../../../docs/architecture.md)), with **every** intermediate `tool-result` re-entering the model's context on the next request.
+In the registry's native presentation, the agent loop advertises every visible capability as a JSON-schema function definition. `ToolRegistry` contributes its schemas to the system-prompt assembly, the assembly's `tools` land on the wire (and in the logged request header), the model invokes one `tool-call` block per step, and at the time of this note the loop dispatched each call through `ctx.tools.execute()` **sequentially** (parallel tool execution was an open TODO then; bounded parallel dispatch has since shipped — the [parallel tool-call note](2026-07-10-parallel-tool-call-execution.md), the rolling pool in [docs/architecture.md](../../../../docs/architecture.md)) — with **every** intermediate `tool-result` re-entering the model's context on the next request.
 
 For multi-step tool work this is token-heavy and serial. The model cannot compose tools — loop over a result set, branch on an intermediate value, fan out, post-process — without a full model round-trip per call, and each round-trip drags the entire intermediate result back into context whether the model needs it or not.
 
@@ -32,7 +32,7 @@ This note owns Code Mode's presentation, composition, isolation, and settlement
 
 **Interaction with `toolOrder`, stated up front:** a configured `systemPrompt.toolOrder` naming native capabilities rejects every assembly under `mode: 'code'`, because those names are outside that mode's wire-validation universe. This is correct behavior, not a bug: a deployment using Code Mode updates its order config or drops it.
 
-**SDK prompt section.** In `'code'` and `'both'`, the lazy `tools:sdk` section in the tool-guidance order band renders TypeScript declarations plus fixed usage instructions for the scope's visible capabilities. It shares lookup and execution visibility, excludes `run_code`, and sorts tools lexicographically for byte-stable output.
+**SDK prompt section.** In `'code'` and `'both'`, the lazy `tools:sdk` section in the tool-guidance order band renders the loaded runtime's language declarations plus fixed usage instructions for the scope's visible capabilities (TypeScript by default; the [language-dispatch note](2026-07-31-code-mode-language-dispatch.md) added Python and the `ctx.codeRuntime.language` renderer table). It shares lookup and execution visibility, excludes `run_code`, and sorts tools lexicographically for byte-stable output.
 
 **Assembly ownership.** `run_code` and `tools:sdk` enter the trusted `system-prompt/assemble` waterfall as normal assembly inputs. A scoped `tools:sdk` section may shadow the global default before dispatch, and a listener may remove or replace either contribution. The waterfall's returned assembly is final, so whoever changes these inputs owns preserving a viable Code Mode protocol when the deployment expects Code Mode to remain usable; no restoration pass overrides deliberate composition.
 
@@ -48,7 +48,7 @@ Under `'code'` and `'both'` the registry owns `run_code` as a reserved presentat
 
 **Sub-call contexts are deferred through the parent.** Injecting inside `run_code` would break parent call/result adjacency, so `ToolRunContext.deferContext()` collects every sub-result `additionalContexts` entry in dispatch order. The registry carries that array even when the program later throws, and the loop appends each entry only after the outer result and every sibling result in the step. An outer post-execute block discards tool-deferred entries and exposes only contexts explicitly attached by the blocking decision.
 
-**Concurrency is serialized.** Each run owns a dispatch queue, so even `Promise.all` executes tool calls in submission order. Settlement abandons queued calls that have not started. Parallelism requires per-tool concurrency-safety metadata.
+**Concurrency is bounded, not serialized.** Each run owns a dispatch queue that starts calls strictly in submission order and classifies each one through `registry.executionMode`, the same fail-closed `isConcurrencySafe` contract the native loop uses. Consecutive parallel-classified calls overlap up to `maxParallelSubCalls` (default 10; `1` restores serial dispatch); an exclusive call drains the pool and runs alone. Settlement abandons queued calls that have not started. This note shipped the serialized placeholder; the [live-parallel Agent Note](2026-07-26-code-mode-live-parallel-dispatch.md) owns the scheduler that replaced it.
 
 **Presentation.** `run_code`'s render intent is decided here per the [render-intent Agent Note](../architecture/2026-07-02-tool-render-intent-union.md): `presentCall` creates a `generic` card with `kind: 'execute'`, the program text as its title, and the same program text as `rawInput`; `run_code` intentionally declares no `presentResult`, so the TUI and host/client runtime (Web) complete that card through their generic raw-content fallback using the final durable `tool/result.content`, including captured logs plus the returned value, failure, or post-policy spill preview. This is not a `terminal` card: that card's semantics are "a shell command in a working directory", which a program is not. See the [result-card completeness note](../../archived/bug-fix/2026-07-20-code-mode-result-card-completeness.md).
 
@@ -64,7 +64,7 @@ Each sub-dispatch appends a log-only `tool/code-dispatch-start` event at pool en
 - `CodeBindingNamespace = { global: string; functions: Record<string, (args: unknown) => Promise<CodeJsonValue>>; errorClass?: { name: string; memberNameProperty: string } }` — the runtime exposes each namespace as a global object of async functions inside the program; the optional descriptor asks the runtime to inject a real program-visible rejection class without teaching the seam consumer-specific names. `CodeJsonValue` is this dependency-light seam's structural lossless-JSON type, so binding arguments and resolutions cross the implementation's serialization boundary whole.
 - `CodeRunResult = { value?: CodeJsonValue; logs: string[]; error?: CodeRunFailure }` — program execution outcomes resolve as the `error` field. `run()` may reject only for caller/seam misuse (for example a duplicate binding namespace); consumers still contain a non-conforming backend rejection at their own error boundary.
 - `CodeRunFailure = { kind: 'exception' | 'timeout' | 'abort' | 'worker-exit' | 'invalid-output' | 'output-limit'; message: string }` — orthogonal outcomes reported independently per [defensive patterns](../../../../docs/defensive-patterns.md); a timed-out run is not an exception, an abort is not a timeout, a lossy completion is not an overflow, and a substrate exit is none of them.
-- Two readonly backend descriptors, informational not gating: `language` (what the program must be written in — `'typescript'` for the shipped backend; a Python backend would say so, and pair with its own SDK generator on the presentation side) and `isolation` (`'worker-thread'` for the shipped backend; `'process'`, `'container'`, … for future ones). `dsh-tools` requires `language === 'typescript'` in the MVP — its codegen emits TS — and fails the assembly loudly otherwise, the same misconfiguration idiom as `toolOrder` violations (as when `mode` is non-native with no `ctx.codeRuntime` loaded at all).
+- Two readonly backend descriptors, informational not gating: `language` (what the program must be written in — `'typescript'` for the first backend; a Python backend says `'python'` and pairs with its own SDK generator on the presentation side) and `isolation` (`'worker-thread'` for the shipped backend; `'process'`, `'container'`, … for future ones). `dsh-tools` accepts any `language` with a registered SDK renderer and `run_code` flavor (TypeScript and Python ship; see the [language-dispatch note](2026-07-31-code-mode-language-dispatch.md)) and fails the assembly loudly otherwise, the same misconfiguration idiom as `toolOrder` violations (as when `mode` is non-native with no `ctx.codeRuntime` loaded at all).
 
 Requests contain every runtime input; implementations own validated timeout and cap defaults. The registry looks up the optional runtime only when Code Mode is assembled, so native mode does not depend on one. Missing or language-incompatible runtimes fail loudly. Alternate substrates or languages can replace the implementation behind the same seam, paired with the appropriate SDK generator.
 
@@ -85,11 +85,11 @@ The worker runtime provides containment, not a security boundary: model code can
 
 ### What the model sees
 
-The SDK instructs the model to write an async erasable-TypeScript body, call tools through `await tools.name(args)`, catch rejected tool calls when needed, and return or log only the output that should re-enter context. Calls remain sequential even under `Promise.all`. The declaration prefix can be as large as native schemas, especially in `'both'`, but remains stable for provider caching.
+The SDK instructs the model to write an async body in the loaded runtime's language (an erasable-TypeScript body by default; a Python `async` body under a Python runtime — see the [language-dispatch note](2026-07-31-code-mode-language-dispatch.md)), call tools through `await tools.name(args)`, catch rejected tool calls when needed, and return or log only the output that should re-enter context. Both flavors state the same contract in their own primitive: independent read-only calls MAY overlap under `Promise.all` (TypeScript) or `asyncio.gather` (Python), mutating calls run alone in submission order, and dependent work sequences with `await`. The declaration prefix can be as large as native schemas, especially in `'both'`, but remains stable for provider caching.
 
 ## Consequences
 
-Deployments switching to `'code'` must update any native-only `toolOrder`. Assembly listeners own the integrity of any rewritten protocol surface. Sub-dispatch remains serialized, while per-call contexts retain their source, envelope, and metadata through the outer result.
+Deployments switching to `'code'` must update any native-only `toolOrder`. Assembly listeners own the integrity of any rewritten protocol surface. Sub-dispatch starts in submission order under a bounded overlap pool, while per-call contexts retain their source, envelope, and metadata through the outer result.
 
 ## Testing
 
@@ -106,7 +106,7 @@ Deployments switching to `'code'` must update any native-only `toolOrder`. Assem
 
 **Result elision / summarization over native tool-calling.** Addresses only the context-bloat half of the problem: trimming old `tool-result`s is cheap to add as a logged surface replacement under reconstructable requests, but still pays one model round-trip per call and cannot express loops, branches, or joins. Complementary, not competing; it can layer under Code Mode for residual native calls.
 
-**Parallel native dispatch in the loop.** The other answer to round-trip cost; still valid future work (the open TODO), still blocked on concurrency-safety metadata, and still no composition — it parallelizes calls the model already decided on in one step. Code Mode's serialized-queue decision keeps the two compatible: when the metadata lands, both native parallel dispatch and per-tool binding parallelism unlock together.
+**Parallel native dispatch in the loop.** The other answer to round-trip cost at decision time; it was blocked on concurrency-safety metadata and offers no composition either way — it parallelizes calls the model already decided on in one step. Code Mode's queue decision kept the two compatible, and that is how it played out: the metadata landed as `isConcurrencySafe` (the [parallel tool-call note](2026-07-10-parallel-tool-call-execution.md)), and native rolling-pool dispatch and per-tool binding parallelism unlocked on the same classifier.
 
 **Always-exclusive (Cloudflare-faithful, no mode).** Rejected for this SDK's primary consumer: a coding agent's bread-and-butter single calls (`bash`, `read`, `edit`) are already ideal as native calls, and forcing every edit through a program taxes the common case. The mode config keeps the faithful form (`'code'`) one line away without imposing it.
 
@@ -128,6 +128,6 @@ Deployments switching to `'code'` must update any native-only `toolOrder`. Assem
 
 **Large lossless JSON values can exhaust memory.** Tool bindings snapshot lossless JSON before dispatch and return canonical JSON resolutions whole. The runtime validates both sides of the worker port and applies no per-binding byte cap; structured-clone cost and process or worker memory are the practical bounds. The combined outer-output ledger for logs, the completion value, and a failure diagnostic is the only byte-capped boundary.
 
-**Serialized-only sub-dispatch.** `Promise.all` gains no wall-clock parallelism yet, only fewer round-trips; models may over-expect. The instructions state it; lifting it is tied to the same concurrency-safety metadata the native parallel-dispatch TODO needs.
+**Sub-dispatch overlap is bounded by tool safety claims, not by the caller.** A program's `Promise.all` or `asyncio.gather` buys wall-clock parallelism only across calls the tool itself classifies concurrency-safe; a run of exclusive calls still costs its round-trips in sequence, and models may over-expect. Both flavors' SDK instructions state the real contract. This note shipped the serialized placeholder that made the risk absolute; the [live-parallel Agent Note](2026-07-26-code-mode-live-parallel-dispatch.md) owns the scheduler and its overlap cap.
 
 **Budget metering reads the event loop, not a flag.** Busy-time polling (`eventLoopUtilization()`) is coarser than an exact CPU meter — a budget expires up to one poll interval late — and its correctness claim ("a pending dispatch cannot pause it") is load-bearing against a hostile program. Both sides are unit-tested (hot loop with a pending decoy dispatch dies at `computeMs`; idle-on-slow-binding survives to `maxWallMs`), and the poll interval is an internal constant, not config — nothing a deployment could mis-tune into a bypass. `maxWallMs` is config, and it reaches `setTimeout`, which clamps a delay above `MAX_TIMER_DELAY_MS` (2^31-1 ms) to 1 ms; a positivity check alone therefore accepts a 25-day ceiling that expires on the first tick and times out every run. The worker runtime range-checks the field at load for that reason. `computeMs` needs no upper bound because it is compared against measured utilization instead of being handed to a timer.

.agents/notes/implemented/feature/2026-06-15-code-mode.zh.md

@@ -6,7 +6,7 @@ Status: implemented
 
 ## 问题
 
-在注册表的原生呈现方式下，agent loop（智能体循环）将每个可见能力以 JSON Schema 函数定义的形式通告给模型。`ToolRegistry` 将其 schema 贡献给系统提示词组装，组装结果中的 `tools` 落到协议格式（wire format）上（也记录在请求头日志中），模型每步调用一个 `tool-call` 块，循环通过 `ctx.tools.execute()` **逐个**分发每次调用（并行工具执行是 `dsh-tools` 和 [docs/architecture.md](../../../../docs/architecture.md) 中明确标注的 open TODO），且**每一个**中间 `tool-result` 都会在下一次请求时重新进入模型上下文。
+在注册表的原生呈现方式下，agent loop（智能体循环）将每个可见能力以 JSON Schema 函数定义的形式通告给模型。`ToolRegistry` 将其 schema 贡献给系统提示词组装，组装结果中的 `tools` 落到协议格式（wire format）上（也记录在请求头日志中），模型每步调用一个 `tool-call` 块，而在本 note 写作时，循环通过 `ctx.tools.execute()` **逐个**分发每次调用（并行工具执行当时还是 open TODO；此后有界的并行分发已经交付——见[并行工具调用 note](2026-07-10-parallel-tool-call-execution.md)，以及 [docs/architecture.md](../../../../docs/architecture.md) 中的 rolling pool）——且**每一个**中间 `tool-result` 都会在下一次请求时重新进入模型上下文。
 
 对于多步工具操作，这种方式 token 开销大且串行。模型无法组合工具——遍历结果集、根据中间值分支、扇出、后处理——每次调用都需要一次完整的模型往返，而每次往返都会把完整的中间结果拖回上下文，不管模型是否需要。
 
@@ -32,7 +32,7 @@ Cloudflare 的 [Code Mode](https://blog.cloudflare.com/code-mode/) 提出了一
 
 **与 `toolOrder` 的交互，预先说明：** 如果配置的 `systemPrompt.toolOrder` 引用了原生能力名称，在 `mode: 'code'` 下会拒绝所有组装，因为那些名称不在该模式的协议校验范围内。这是正确行为而非 bug：使用 Code Mode 的部署需要更新其 order 配置或移除它。
 
-**SDK 提示词段。** 在 `'code'` 和 `'both'` 下，tool-guidance order band 中的惰性 `tools:sdk` 段为当前 scope 的可见能力渲染 TypeScript 声明加固定的使用说明。它共享查找和执行可见性，排除 `run_code`，并按字典序排列工具以获得字节稳定的输出。
+**SDK 提示词段。** 在 `'code'` 和 `'both'` 下，tool-guidance order band 中的惰性 `tools:sdk` 段为当前 scope 的可见能力渲染所加载运行时语言的声明加固定的使用说明（默认 TypeScript；[语言分发 note](2026-07-31-code-mode-language-dispatch.md) 加入了 Python 与按 `ctx.codeRuntime.language` 选择的渲染器表）。它共享查找和执行可见性，排除 `run_code`，并按字典序排列工具以获得字节稳定的输出。
 
 **组装所有权。** `run_code` 和 `tools:sdk` 作为正常的组装输入进入受信任的 `system-prompt/assemble` waterfall。一个 scoped 的 `tools:sdk` 段可以在分发前遮蔽全局默认值，监听器也可以移除或替换任一贡献。waterfall 返回的组装结果是最终的，因此修改这些输入的人有责任在部署期望 Code Mode 可用时保持协议面的完整性；没有恢复 pass 会覆盖有意的组合。
 
@@ -48,7 +48,7 @@ Cloudflare 的 [Code Mode](https://blog.cloudflare.com/code-mode/) 提出了一
 
 **子调用上下文通过父调用延后。** 在 `run_code` 内部注入会破坏父调用/结果的相邻性，因此 `ToolRunContext.deferContext()` 按分发顺序收集每个子结果的 `additionalContexts` 条目。即使程序后来抛出异常，注册表仍携带该数组；循环只在外层结果与步骤中所有兄弟结果之后追加每个条目。外层 post-execute 阻止会丢弃工具延后的条目，只暴露阻止 decision 显式附加的上下文。
 
-**并发被序列化。** 每次 run 拥有一个分发队列，因此即使 `Promise.all` 也按提交顺序执行工具调用。结算时放弃尚未开始的排队调用。并行化需要每个工具的并发安全元数据。
+**并发是有界的，而非被序列化。** 每次 run 拥有一个分发队列，严格按提交顺序启动调用，并通过 `registry.executionMode` 对每个调用分类——与原生循环所用的 fail-closed `isConcurrencySafe` 契约相同。连续的 parallel 类调用最多重叠 `maxParallelSubCalls` 个（默认 10；设为 `1` 恢复串行分发）；exclusive 类调用会排空池并单独运行。结算时放弃尚未开始的排队调用。本 note 交付的是被序列化的占位实现；取代它的调度器由[实时并行 Agent Note](2026-07-26-code-mode-live-parallel-dispatch.md) 负责。
 
 **呈现。** `run_code` 的 render intent 按[呈现意图 Agent Note](../architecture/2026-07-02-tool-render-intent-union.md)在此决定：`presentCall` 创建一个 `generic` 卡片，`kind: 'execute'`，以程序文本作为标题，并将同一程序文本作为 `rawInput`；`run_code` 有意不声明 `presentResult`，因此 TUI 和宿主/客户端运行时（Web）会通过通用原始内容回退机制，使用最终持久化的 `tool/result.content` 补全该卡片，其中包括捕获的日志，以及返回值、失败信息或 post-policy 输出落盘预览。这不是 `terminal` 卡片：该卡片的语义是「工作目录中的 shell 命令」，程序不是。参见[结果卡片完整性说明](../../archived/bug-fix/2026-07-20-code-mode-result-card-completeness.md)。
 
@@ -64,7 +64,7 @@ Cloudflare 的 [Code Mode](https://blog.cloudflare.com/code-mode/) 提出了一
 - `CodeBindingNamespace = { global: string; functions: Record<string, (args: unknown) => Promise<CodeJsonValue>>; errorClass?: { name: string; memberNameProperty: string } }`——运行时将每个命名空间作为程序内部的全局异步函数对象暴露；可选描述符要求运行时注入真正的、程序可见的 reject 类，而无需让 seam 获知消费方专用名称。`CodeJsonValue` 是这个低依赖 seam 的结构化无损 JSON 类型，因此绑定参数与返回值可以完整跨越实现的序列化边界。
 - `CodeRunResult = { value?: CodeJsonValue; logs: string[]; error?: CodeRunFailure }`——程序执行失败时，执行 promise 仍会 fulfill，并通过 `error` 字段返回失败结果。只有调用方/seam 误用（例如重复的绑定命名空间）时，`run()` 才会 reject；消费方仍在自己的错误边界处理不合规后端的拒绝。
 - `CodeRunFailure = { kind: 'exception' | 'timeout' | 'abort' | 'worker-exit' | 'invalid-output' | 'output-limit'; message: string }`——按[防御性模式](../../../../docs/defensive-patterns.md)独立报告的正交结果；超时的 run 不是异常，abort 不是超时，有损完成值不是溢出，基底退出也与上述情况相互独立。
-- 两个只读的后端描述符，仅供信息参考而非门禁判定：`language`（程序必须使用的语言——交付的后端为 `'typescript'`；Python 后端会声明自己，并在呈现侧配对自己的 SDK 生成器）和 `isolation`（交付的后端为 `'worker-thread'`；未来可为 `'process'`、`'container'` 等）。`dsh-tools` 在 MVP 中要求 `language === 'typescript'`——其代码生成输出 TS——否则组装会大声失败，与 `toolOrder` 违规时的配置错误惯用法相同（如 `mode` 为非 native 但根本没有加载 `ctx.codeRuntime`）。
+- 两个只读的后端描述符，仅供信息参考而非门禁判定：`language`（程序必须使用的语言——首个后端为 `'typescript'`；Python 后端声明 `'python'`，并在呈现侧配对自己的 SDK 生成器）和 `isolation`（交付的后端为 `'worker-thread'`；未来可为 `'process'`、`'container'` 等）。`dsh-tools` 接受任何注册了 SDK 渲染器与 `run_code` flavor 的 `language`（TypeScript 与 Python 已交付；见[语言分发 note](2026-07-31-code-mode-language-dispatch.md)），否则组装会大声失败，与 `toolOrder` 违规时的配置错误惯用法相同（如 `mode` 为非 native 但根本没有加载 `ctx.codeRuntime`）。
 
 请求包含所有运行时输入；实现方拥有经校验的超时和上限默认值。注册表仅在组装 Code Mode 时查找可选的运行时，因此 native 模式不依赖它。缺失或语言不兼容的运行时会大声失败。替代基底或语言可以在同一 seam 背后替换实现，配对相应的 SDK 生成器。
 
@@ -85,11 +85,11 @@ worker 运行时只能约束程序的运行，而不构成安全边界：模型
 
 ### 模型看到的内容
 
-SDK 指示模型编写一个异步的可擦除 TypeScript 函数体，通过 `await tools.name(args)` 调用工具，在需要时捕获被拒绝的工具调用，并仅 return 或 log 应重新进入上下文的输出。即使在 `Promise.all` 下调用仍保持顺序。声明前缀可能与原生 schema 一样大，尤其在 `'both'` 下，但对提供方缓存保持稳定。
+SDK 指示模型编写一个所加载运行时语言的异步函数体（默认可擦除 TypeScript；Python 运行时下为 Python `async` 函数体——见[语言分发 note](2026-07-31-code-mode-language-dispatch.md)），通过 `await tools.name(args)` 调用工具，在需要时捕获被拒绝的工具调用，并仅 return 或 log 应重新进入上下文的输出。两种 flavor 用各自的原语陈述同一契约：相互独立的只读调用可以在 `Promise.all`（TypeScript）或 `asyncio.gather`（Python）下重叠，有副作用的调用按提交顺序单独运行，有依赖的工作用 `await` 排序。声明前缀可能与原生 schema 一样大，尤其在 `'both'` 下，但对提供方缓存保持稳定。
 
 ## 后果
 
-切换到 `'code'` 的部署必须更新任何仅限 native 的 `toolOrder`。组装监听器有责任维护任何被重写的协议面的完整性。子分发保持序列化，而每次调用的上下文会通过外层结果保留其 source、信封与元数据。
+切换到 `'code'` 的部署必须更新任何仅限 native 的 `toolOrder`。组装监听器有责任维护任何被重写的协议面的完整性。子分发在有界的重叠池下按提交顺序启动，而每次调用的上下文会通过外层结果保留其 source、信封与元数据。
 
 ## 测试
 
@@ -106,7 +106,7 @@ SDK 指示模型编写一个异步的可擦除 TypeScript 函数体，通过 `aw
 
 **在原生工具调用上做结果省略/摘要。** 仅解决问题中上下文膨胀这一半：裁剪旧 `tool-result` 作为可重建请求下的日志化表面替换成本低，但仍需每次调用一次模型往返，且无法表达循环、分支或汇合。互补而非竞争；它可以在 Code Mode 下为残余的原生调用分层。
 
-**循环中的并行原生分发。** 往返成本的另一个答案；仍是有效的未来工作（open TODO），仍被并发安全元数据阻塞，且仍无组合能力——它并行化的是模型在一步中已经决定的调用。Code Mode 的序列化队列决策保持两者兼容：当元数据就绪时，原生并行分发和每工具绑定并行化一起解锁。
+**循环中的并行原生分发。** 决策当时对往返成本的另一个答案；它被并发安全元数据阻塞，且无论如何都不提供组合能力——它并行化的是模型在一步中已经决定的调用。Code Mode 的队列决策保持了两者兼容，后续也正是这样落地的：元数据以 `isConcurrencySafe` 的形式就绪（见[并行工具调用 note](2026-07-10-parallel-tool-call-execution.md)），原生 rolling-pool 分发与每工具绑定并行化基于同一个分类器一起解锁。
 
 **始终排他（忠于 Cloudflare，无模式）。** 否决，因为本 SDK 的主要消费方是编码 agent：其日常的单次调用（`bash`、`read`、`edit`）作为原生调用已经是最优的，强制每次编辑都通过程序会给常见场景增加负担。mode 配置让忠实形式（`'code'`）只需一行配置即可启用，而不强加于人。
 
@@ -128,6 +128,6 @@ SDK 指示模型编写一个异步的可擦除 TypeScript 函数体，通过 `aw
 
 **大型无损 JSON 值可能耗尽内存。** 工具绑定会在分发前对无损 JSON 创建快照，并完整返回规范 JSON 返回值。运行时会校验 worker 端口两侧，但不对单次绑定设置字节数上限；结构化克隆成本以及进程或 worker 内存构成实际边界。只有包含日志、完成值和失败诊断的组合外层输出账本受字节数上限约束。
 
-**仅序列化的子分发。** `Promise.all` 尚未获得挂钟并行性，仅减少往返次数；模型可能过度期望。说明中已声明；解除此限制与原生并行分发 TODO 所需的并发安全元数据绑定。
+**子分发的重叠由工具自身的安全声明限定，而非由调用方决定。** 程序里的 `Promise.all` 或 `asyncio.gather` 只在工具自己分类为并发安全的调用之间换来挂钟并行性；一串 exclusive 调用仍要按顺序付出各自的往返开销，模型可能过度期望。两种 flavor 的 SDK 说明都陈述了真实契约。本 note 交付的是使该风险绝对化的序列化占位实现；调度器及其重叠上限由[实时并行 Agent Note](2026-07-26-code-mode-live-parallel-dispatch.md) 负责。
 
 **预算计量读取事件循环，而非 flag。** 忙碌时间轮询（`eventLoopUtilization()`）比精确 CPU 计量更粗糙——预算到期最多延迟一个轮询间隔——且其正确性声明（「pending 的分发不能暂停它」）是抵御恶意程序的关键。两种情况均有单元测试（带 pending 诱饵分发的热循环会在耗尽 `computeMs` 预算时终止；等待慢速绑定的空闲程序则会持续运行至 `maxWallMs`），轮询间隔是内部常量而非配置——部署无法将其误调为绕过手段。`maxWallMs` 是配置项，且会传入 `setTimeout`，后者会把超过 `MAX_TIMER_DELAY_MS`（2^31-1 ms）的延迟夹到 1 ms；因此仅有正数校验会放行一个 25 天的上限，它在第一个 tick 就到期，使每次运行都超时。worker 运行时正因如此在加载时对该字段做范围校验。`computeMs` 不需要上界，因为它对照的是实测占用率，而不是交给定时器。

.agents/notes/implemented/feature/2026-06-21-subagent-capability-seam.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/2026-06-21-subagent-capability-seam.md
-2026-06-21-subagent-capability-seam.md: 043092884731c403a11b71ef8b5a410e9e5af7e0
-2026-06-21-subagent-capability-seam.zh.md: 6a4a5798199ca7d6d7c011668a319d65a0553208
+2026-06-21-subagent-capability-seam.md: 752e639b09ea2ac0ba19841ddfe38b15b44d22e9
+2026-06-21-subagent-capability-seam.zh.md: 6009aeda6773357a4217f8956fb510c2116bbe3f

.agents/notes/implemented/feature/2026-06-21-subagent-capability-seam.md

@@ -4,7 +4,7 @@ Status: implemented
 
 English | [中文](2026-06-21-subagent-capability-seam.zh.md)
 
-> The full seam is shipped: the `dsh-subagent` interface and `dsh-tool-subagent` consumer; the two in-process backends (`dsh-subagent-spawn`, `dsh-subagent-fork`); the nested-agent snapshot infrastructure ([per-session snapshot replay](../testing/2026-06-22-subagent-snapshot-replay.md)); and the out-of-process `dsh-subagent-acp` backend ([its Agent Note](2026-06-22-acp-subagent-backend.md)).
+> The full seam is shipped: the `dsh-subagent` interface and `dsh-tool-subagent` consumer; the two in-process backends (`dsh-subagent-spawn`, `dsh-subagent-fork`); the nested-agent snapshot infrastructure ([per-session snapshot replay](../testing/2026-06-22-subagent-snapshot-replay.md)); and the out-of-process ACP, Codex, and Claude Code backends ([ACP Agent Note](2026-06-22-acp-subagent-backend.md), [product-provider Agent Note](2026-08-04-claude-code-and-codex-subagent-backends.md)).
 
 ## Problem
 
@@ -14,7 +14,8 @@ The distinctive requirement — the one that shapes the whole design — is that
 
 - **in-process** — a child concrete `Agent` on the same `Context` (the cheapest, and nearly free given the existing agent factory);
 - **ACP** — act as an ACP *client* driving another agent process (which can be another instance of ourselves);
-- later: **A2A**, the **Codex app-server**, and the **Claude Code Agent SDK** — each the same out-of-process "start a child, prompt it, stream updates, cancel" shape as the ACP backend.
+- **Codex app-server and Claude Code Agent SDK** — current one-shot siblings that apply the same named-provider seam to official product processes ([product-provider Agent Note](2026-08-04-claude-code-and-codex-subagent-backends.md));
+- later: **A2A** using the same out-of-process "start a child, prompt it, settle, cancel" shape.
 
 ## Alternatives considered
 
@@ -34,6 +35,8 @@ A new package group `packages/subagent/`:
 | `@deepseek-ai/dsh-subagent-spawn` | implementation: a fresh in-process child via `ctx.agents.create` |
 | `@deepseek-ai/dsh-subagent-fork` | implementation: an in-process child seeded with a snapshot of the parent's log |
 | `@deepseek-ai/dsh-subagent-acp` | implementation: an ACP client driving a configured child process |
+| `@deepseek-ai/dsh-subagent-codex` | implementation: a one-shot official Codex app-server process |
+| `@deepseek-ai/dsh-subagent-claude-code` | implementation: a one-shot official Claude Code process through the Agent SDK |
 | `@deepseek-ai/dsh-tool-subagent` | consumer: the model-facing `subagent` tool over `ctx.subagents` |
 
 ### The primitive: async `start → SubagentRun`
@@ -51,7 +54,7 @@ Fresh and forked children are separate providers, not a request flag. `dsh-subag
 
 ### Child isolation and the parent log
 
-Each subagent runs in its **own `Session`** (own id, `parentSession` lineage), persisted independently. The parent's log records only the spawn `tool/call` and its `tool/result` (the child's final output) — the child's internal steps and tool calls stay in the child's own session, never injected into the parent log. This is the only design that is identical across transports: an ACP child's internal events physically cannot be injected into our parent log, so making in-process behave the same keeps the seam transport-agnostic.
+Each in-process subagent runs in its **own `Session`** (own id, `parentSession` lineage), persisted independently. Remote ACP and one-shot product providers instead mint a parent-scoped lifecycle id and expose no local `Agent` or child `Session`; their internal state remains in the remote process. Across both forms, the parent's log records only the spawn `tool/call` and its `tool/result` (the child's final output), while child steps and tool calls remain outside the parent log.
 
 ### Synchronous collect (first cut)
 

.agents/notes/implemented/feature/2026-06-21-subagent-capability-seam.zh.md

@@ -4,7 +4,7 @@ Status: implemented
 
 [English](2026-06-21-subagent-capability-seam.md) | 中文
 
-> 完整 seam 已交付：`dsh-subagent` 接口与 `dsh-tool-subagent` 消费方；两个进程内后端（`dsh-subagent-spawn`、`dsh-subagent-fork`）；嵌套 agent 快照基础设施（[逐会话快照回放](../testing/2026-06-22-subagent-snapshot-replay.md)）；以及进程外后端 `dsh-subagent-acp`（[其 Agent Note](2026-06-22-acp-subagent-backend.md)）。
+> 完整 seam 已交付：`dsh-subagent` 接口与 `dsh-tool-subagent` 消费方；两个进程内后端（`dsh-subagent-spawn`、`dsh-subagent-fork`）；嵌套 agent 快照基础设施（[逐会话快照回放](../testing/2026-06-22-subagent-snapshot-replay.md)）；以及进程外的 ACP、Codex 与 Claude Code 后端（[ACP Agent Note](2026-06-22-acp-subagent-backend.md)、[产品提供方 Agent Note](2026-08-04-claude-code-and-codex-subagent-backends.md)）。
 
 ## 问题
 
@@ -14,7 +14,8 @@ harness 有一个长期搁置的 seam 用于 **subagent**：一个 agent（智
 
 - **进程内**：在同一个 `Context` 上创建一个具体的子 `Agent`（最廉价，且鉴于现有 agent 工厂几乎零成本）；
 - **ACP**：作为 ACP *客户端*驱动另一个 agent 进程（可以是自身的另一个实例）；
-- 后续：**A2A**、**Codex app-server** 与 **Claude Code Agent SDK**——每种都与 ACP 后端相同的进程外形状：「启动子 agent、发送提示词、流式接收更新、取消」。
+- **Codex app-server 与 Claude Code Agent SDK**：当前的一次性兄弟提供方，将同一个命名提供方 seam 应用于官方产品进程（[产品提供方 Agent Note](2026-08-04-claude-code-and-codex-subagent-backends.md)）；
+- 后续：**A2A**，采用同样的进程外形态：「启动子 agent、发送提示词、结算、取消」。
 
 ## 曾考虑的替代方案
 
@@ -34,6 +35,8 @@ bash seam（[能力 seam](../architecture/2026-06-13-capability-seams.md)）在
 | `@deepseek-ai/dsh-subagent-spawn` | 实现：通过 `ctx.agents.create` 创建全新的进程内子 agent |
 | `@deepseek-ai/dsh-subagent-fork` | 实现：用父 agent 日志快照初始化的进程内子 agent |
 | `@deepseek-ai/dsh-subagent-acp` | 实现：作为 ACP 客户端驱动已配置的子进程 |
+| `@deepseek-ai/dsh-subagent-codex` | 实现：一次性官方 Codex app-server 进程 |
+| `@deepseek-ai/dsh-subagent-claude-code` | 实现：通过 Agent SDK 运行的一次性官方 Claude Code 进程 |
 | `@deepseek-ai/dsh-tool-subagent` | 消费方：基于 `ctx.subagents` 的面向模型的 `subagent` 工具 |
 
 ### 原语：异步 `start → SubagentRun`
@@ -51,7 +54,7 @@ bash seam（[能力 seam](../architecture/2026-06-13-capability-seams.md)）在
 
 ### 子 agent 隔离与父日志
 
-每个 subagent 运行在**自己的 `Session`** 中（独立 id、`parentSession` 谱系），独立持久化。父日志仅记录 spawn `tool/call` 及其 `tool/result`（子 agent 的最终输出）——子 agent 的内部步骤和工具调用留在子 agent 自己的会话中，绝不注入父日志。这是唯一在所有传输方式下行为一致的设计：ACP 子 agent 的内部事件在物理上无法注入我们的父日志，因此让进程内行为保持一致，使 seam 真正与传输方式无关。
+每个进程内 subagent 运行在**自己的 `Session`** 中（独立 id、`parentSession` 谱系），独立持久化。远端 ACP 和一次性产品提供方则会生成一个父级作用域的生命周期 id，且不暴露本地 `Agent` 或子 `Session`；其内部状态留在远端进程中。两种形式下，父日志都仅记录 spawn `tool/call` 及其 `tool/result`（子 agent 的最终输出），而子 agent 的步骤和工具调用均留在父日志之外。
 
 ### 同步收集（首版）
 

.agents/notes/implemented/feature/2026-06-22-acp-subagent-backend.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/2026-06-22-acp-subagent-backend.md
-2026-06-22-acp-subagent-backend.md: 5f12aa1c08d4f4cfaa35f2f7f4b09ad341c3eae8
-2026-06-22-acp-subagent-backend.zh.md: ba9d4255723367ab4afc5ab87e056f12f5c3a285
+2026-06-22-acp-subagent-backend.md: c994ebfa69649bb9e79d3aa389a0e13c178131c7
+2026-06-22-acp-subagent-backend.zh.md: 62d631e95b5d1f299bed1fd00353b5ac9349b9a0

.agents/notes/implemented/feature/2026-06-22-acp-subagent-backend.md

@@ -57,6 +57,6 @@ Persistent-process pooling (reuse a warm child across runs) is a performance opt
 
 Every run pays a fresh subprocess (spawn + `initialize` + `newSession`). The parent surfaces only the child's final answer: `session/update` thoughts and tool-call cards are consumed and dropped, and permission prompts never reach a human — the configured policy answers them. The child's environment is credential-scrubbed by default, so its own model key is supplied explicitly via `config.env`.
 
-## Future providers
+## Product-provider siblings
 
-The same out-of-process spawn/prompt/stream/cancel shape generalizes to other transports named in the seam Agent Note — A2A, the Codex app-server, and the Claude Code Agent SDK — each a sibling provider registered by name. The ACP backend is the proof that the seam supports the boundary; those are mechanically similar.
+The [Codex app-server and Claude Code Agent SDK providers](2026-08-04-claude-code-and-codex-subagent-backends.md) apply the same out-of-process spawn/prompt/settle/cancel boundary as siblings registered by name. A2A remains a future sibling transport; the ACP backend proves that the common seam supports this boundary without owning product-private protocols.

.agents/notes/implemented/feature/2026-06-22-acp-subagent-backend.zh.md

@@ -57,6 +57,6 @@ ACP `StopReason` → harness `SubagentStopReason`：`end_turn`→`completed`、`
 
 每次运行都要付出一个全新子进程的代价（spawn + `initialize` + `newSession`）。父进程仅暴露子 agent 的最终回答：`session/update` 中的思考和工具调用卡片被消费后丢弃，权限提示从不到达人类——由配置的策略应答。子进程环境默认经过凭证清洗，因此其自身的模型密钥需通过 `config.env` 显式提供。
 
-## 后续提供方
+## 兄弟产品提供方
 
-同样的进程外启动/提示词/流式输出/取消形态可泛化到 seam Agent Note 中列出的其他传输方式——A2A、Codex app-server 和 Claude Code Agent SDK——每个都是按名称注册的兄弟提供方。ACP 后端证明了 seam 支持跨进程边界；其余在机制上类似。
+[Codex app-server 与 Claude Code Agent SDK 提供方](2026-08-04-claude-code-and-codex-subagent-backends.md)作为按名称注册的兄弟提供方，采用同样的进程外启动/提示词/结算/取消边界。A2A 仍是未来的兄弟传输方式；ACP 后端证明了通用 seam 能够支持这项边界，而无需负责产品私有协议。

.agents/notes/implemented/feature/2026-06-24-workspace-context.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/2026-06-24-workspace-context.md
-2026-06-24-workspace-context.md: df7ae58f8a42a8c1aac8113a429ef0f50b2fb795
-2026-06-24-workspace-context.zh.md: aecc57a5c63b4c6282260863f4183944c14b5cf8
+2026-06-24-workspace-context.md: 3c21b7599e6bb2dd759e4a040c1b999f1b0516cd
+2026-06-24-workspace-context.zh.md: e913ecb22b4692e1ee7b96d4f5effe2606775cf8

.agents/notes/implemented/feature/2026-06-24-workspace-context.md

@@ -30,9 +30,9 @@ The user-global file is fixed at `$DSH_HOME/AGENTS.md`, is not affected by eithe
 
 At the first `agent/pre-step` of an agent-loop instance, the plugin composes one sourced user-role baseline. When the downstream decision enters a nonempty first-step batch, the plugin folds the baseline into that final batch right after the claimed prompt, so it becomes durable with the direct prompt and reaches the first request. Rejection or an empty first-step decision leaves the baseline in the next-step inbox for a later wakeup. The plugin loads the user-global file first, then finds the project root by walking upward from `agent.session.header.cwd` to a configured root marker (default `.git`), then loads the configured candidates from each directory from the root to the cwd. A `.git` file and a `.git` directory are both valid markers, covering linked worktrees and submodules. Without a marker, the cwd itself is the root.
 
-The baseline becomes a durable `user/message` with a typed `workspace-instructions` source. Its `baseline: true` marker distinguishes the complete startup or resume baseline from later deltas, and its change list persists the included scopes and content digests. If a previously queued workspace baseline is still pending, the plugin removes that exact message and prepends its replacement instead of accumulating duplicates.
+The baseline becomes a durable `user/message` with a typed `workspace-instructions` source. Its `baseline: true` marker distinguishes a complete baseline from later deltas, its `baselineIdentity` records normalized discovery, precedence, project-root, and budget semantics, and its change list persists the included scopes and content digests. If a previously queued workspace baseline is still pending, the plugin removes that exact message and prepends its replacement instead of accumulating duplicates.
 
-A resumed agent creates a new loop instance and injects a baseline composed from current files before its first request. This permits current baseline content on resume without mutating an earlier history event. A resume and a hot plugin remount both face a log that may already hold a baseline; they are told apart by `agent/session-start`, which a startup or resume emits before the first step while a remount attaches to an already-live session and never sees it. A remount retains the existing baseline only when its typed event remains in the current visible surface, and still rebuilds scope and provider-version tracking from current files. If compaction has shadowed that event, the remount injects a current baseline. A resume always re-composes.
+A resumed agent creates a new loop instance over persisted history. At its first `agent/pre-step`, a visible baseline with the current identity remains authoritative while the plugin compares its retained scopes with a complete current rendering. Unchanged and budget-omitted files append nothing; offline additions, edits, removals, and files that leave the retained budget set append `set`, `replace`, or `remove` transitions in the entering batch without mutating or duplicating the original baseline. An incompatible visible baseline is superseded by one complete baseline in current precedence order with explicit replacement language; when no current candidate exists, an explicit empty baseline clears the earlier scopes. A hot plugin remount follows the same rule. If compaction has shadowed the typed baseline, the next entering pre-step composes one complete current baseline and carries it in the same request.
 
 The baseline is a user-role `<system-reminder>` with `Instructions from: <path>` sections and explicit authority and precedence language. This familiar model-facing frame avoids a harness-specific XML vocabulary. Project paths are root-relative and the user-global path is `~/.dsh/AGENTS.md` for the default home or `$DSH_HOME/AGENTS.md` for a configured home. The final rendering boundary escapes a literal `</system-reminder>` anywhere in instruction content or model-visible path, scope, and budget metadata before byte accounting completes. The package README owns the exact current [prompt shape](../../../../packages/context/workspace-context/README.md#prompt-shape).
 
@@ -48,15 +48,15 @@ Shell commands are not discovery triggers. Local bash calls start fresh shells,
 
 ### Duplicate Suppression And Change Detection
 
-Every workspace context event stores versioned metadata with `{ action, scope, path, digest? }`, where `digest` is SHA-1 over the loaded content. Baselines additionally carry `baseline: true`. The model-facing prompt has no HTML comments, hidden markers, or headings that are parsed back into state.
+Every workspace context event stores versioned metadata with `{ action, scope, path, digest? }`, where `digest` is SHA-1 over the loaded content. Complete baselines additionally carry `baseline: true` and `baselineIdentity`. The model-facing prompt has no HTML comments, hidden markers, or headings that are parsed back into state.
 
 At reconciliation time the plugin scans workspace-sourced `user/message` events and derives the latest state for each visible scope. A short per-session pending map begins only after the immutable top-level `tools/result` proves an `additionalContexts` entry survived every post-execute listener, then covers the interval before the loop appends that context to the log. Each entry records the open `{ turn, step }`: an equal durable `user/message` at or after its sequence boundary confirms and removes it, while a matching `step/end` arriving first means the loop discarded its context buffer, so the plugin removes both the pending entry and its version-cache fast path. A nested Code Mode result stages its changes under the parent's opaque execution token so repeated sub-dispatches in one run do not duplicate them; the parent result rolls that provisional state back and commits only contexts retained by outer policy.
 
-An unchanged path and digest is suppressed. A logged removal is a tombstone, so a reappearing candidate becomes a new `set`. Resume works from persisted metadata. If compaction removes an instruction event from the visible surface, that state no longer suppresses a later load, matching the fact that the model can no longer see it. Only changes actually included under the byte budget enter metadata or pending state, so an omitted file remains eligible on a later touch.
+An unchanged path and digest is suppressed. A logged removal is a tombstone, so a reappearing candidate becomes a new `set`. Resume works from persisted metadata: a compatible visible baseline supplies comparison state rather than causing another complete baseline to be appended. If compaction removes an instruction event from the visible surface, that state no longer suppresses a later load, matching the fact that the model can no longer see it. Only changes actually included under the byte budget enter metadata or pending state, so an omitted file remains eligible on a later touch.
 
-The initial baseline's typed changes are comparison state only while its event remains in the visible session surface. A later successful filesystem touch re-adds an unchanged baseline scope after compaction, or appends baseline edits or removals as dynamic messages; it never rewrites the original event. The in-memory scope marker and provider-version cache only select and accelerate probes, so neither can suppress context the model no longer sees. During resumed baseline preparation the plugin also reconciles visible dynamic scopes, so nested changes made while the agent was offline can append an update before the first resumed request.
+The initial baseline's typed changes are comparison state only while its event remains in the visible session surface. Resume retains a compatible baseline and reconciles current baseline and visible dynamic scopes, so changes made while the agent was offline append transitions before the first resumed request. When compaction shadows the baseline event, the next entering pre-step composes the complete current baseline and records it in the same request; a successful filesystem touch can instead re-add an unchanged baseline scope or append baseline edits or removals as dynamic messages. Neither path rewrites the original event. The in-memory scope marker and provider-version cache only select and accelerate probes, so neither can suppress context the model no longer sees.
 
-There is intentionally no watcher. Detection occurs at the next successful structured filesystem touch or resumed baseline preparation. A provider failure produces no removal; absence is only accepted when all configured candidates in that scope were probed successfully.
+There is intentionally no watcher. Detection occurs at the next successful structured filesystem touch, resumed-baseline reconciliation, or entering pre-step that restores a shadowed baseline. A provider failure produces no removal; absence is only accepted when all configured candidates in that scope were probed successfully.
 
 ### Byte Budget And Bounded Reads
 
@@ -82,7 +82,7 @@ Workspace guidance is isolated per session and shared by the demo front doors, W
 
 Repository text remains untrusted input. Lower-authority user-role framing, explicit precedence language, and delimiter escaping reduce risk but do not eliminate prompt injection. Following a candidate symlink to its target widens that surface to off-tree content, so the permission and sandbox layers that confine `ctx.fs` to trusted roots are the boundary that treats workspace files as data rather than authority (the [instruction-symlink follow note](2026-07-21-follow-instruction-symlinks.md) owns the residual risk).
 
-The system is event-driven rather than watch-driven. Edits are not visible at the exact filesystem mutation instant unless that mutation goes through a structured tool; externally changed files are noticed on the next successful structured touch or resume. This keeps the design deterministic and provider-neutral.
+The system is event-driven rather than watch-driven. Edits are not visible at the exact filesystem mutation instant unless that mutation goes through a structured tool; externally changed files are noticed on the next successful structured touch, resume reconciliation, or restoration of a shadowed baseline. This keeps the design deterministic and provider-neutral.
 
 ## Deferred
 

.agents/notes/implemented/feature/2026-06-24-workspace-context.zh.md

@@ -30,9 +30,9 @@ Status: implemented
 
 在 agent loop（智能体循环）实例的第一个 `agent/pre-step`，插件会组合一条带来源的 user 角色基线。当下游决策让非空的第一步批次进入时，插件会将基线折入最终批次、紧随已领取的直接提示词之后，使其与直接提示词一同成为持久记录并抵达第一次请求。reject 或空的第一步决策会将基线留在 next-step inbox，等待后续唤醒。插件先加载用户全局文件，再从 `agent.session.header.cwd` 向上遍历至配置的根标记（默认为 `.git`）以确定项目根目录，随后从根目录至 cwd 的每级目录加载已配置候选项。`.git` 文件与 `.git` 目录都是有效标记，因而能覆盖链接 worktree 和 submodule。找不到标记时，cwd 本身就是根目录。
 
-该基线会成为一条持久 `user/message`，并携带带类型的 `workspace-instructions` 来源。其 `baseline: true` 标记将完整的启动或恢复基线与后续增量区分开来，变更列表则持久保存已纳入的作用域和内容 digest。若先前排队的 workspace 基线仍在等待，插件会删除该确切消息并 prepend 替代值，而不会累积副本。
+该基线会成为一条持久 `user/message`，并携带带类型的 `workspace-instructions` 来源。其 `baseline: true` 标记将完整基线与后续增量区分开来，`baselineIdentity` 记录规范化的发现、优先级、项目根目录和预算语义，变更列表则持久保存已纳入的作用域和内容 digest。若先前排队的 workspace 基线仍在等待，插件会删除该确切消息并 prepend 替代值，而不会累积副本。
 
-恢复 agent 会创建新的循环实例，并在其第一次请求前注入由当前文件组合的基线。这样，恢复时可以使用当前基线内容，而无需修改先前的历史事件。恢复与插件热重挂都会面对日志中可能已存在基线的情况；二者通过 `agent/session-start` 区分：启动或恢复会在第一步前发出该事件，而热重挂附着到一个已存活的会话、永远不会看到它。只有当基线的类型化事件仍在当前可见表层中时，热重挂才保留既有基线，同时仍会根据当前文件重建 scope 与提供方版本跟踪。如果压缩（compaction）已遮蔽该事件，热重挂会注入当前基线。恢复则始终重新组合。
+恢复 agent 会基于持久化历史创建新的 loop 实例。在第一个 `agent/pre-step`，具有当前标识的可见基线仍是权威状态；插件会将其保留的 scope 与当前完整渲染进行比较。未变化和被预算省略的文件不追加任何内容；agent 离线期间新增、编辑、移除或不再属于预算保留集的文件，会在进入步骤的批次中追加 `set`、`replace` 或 `remove` 转换，既不改写也不重复原始基线。不兼容的可见基线会被一条按当前优先级排列的完整基线取代，并以明确措辞说明替换关系；如果当前不存在任何候选文件，一条显式空基线会清除先前的 scope。插件热重挂遵循相同规则。如果压缩（compaction）已遮蔽带类型的基线，下一次进入步骤的 pre-step 会组合一条完整的当前基线，并在同一请求中携带它。
 
 基线是一条 user 角色的 `<system-reminder>`，包含 `Instructions from: <path>` 章节，以及明确的权威性与优先级说明。这种熟悉的模型可见框架避免引入 harness 专用的 XML 词汇。项目路径相对于根目录；使用默认 home 时，用户全局路径为 `~/.dsh/AGENTS.md`，使用已配置 home 时则为 `$DSH_HOME/AGENTS.md`。最终渲染边界会在完成字节核算前，转义指令内容或模型可见的路径、scope 与预算元数据中出现的字面量 `</system-reminder>`。包 README 负责规定当前准确的[提示词形态](../../../../packages/context/workspace-context/README.md#prompt-shape)。
 
@@ -48,15 +48,15 @@ shell 命令不会触发发现。本地 bash 调用会启动全新的 shell，
 
 ### 重复抑制与变更检测
 
-每个工作区上下文事件都会存储带版本的元数据，其形态为 `{ action, scope, path, digest? }`；`digest` 是对已加载内容计算的 SHA-1。基线还会额外携带 `baseline: true`。模型可见提示词中没有 HTML 注释、隐藏标记，也没有会被解析回状态的标题。
+每个工作区上下文事件都会存储带版本的元数据，其形态为 `{ action, scope, path, digest? }`；`digest` 是对已加载内容计算的 SHA-1。完整基线还会额外携带 `baseline: true` 和 `baselineIdentity`。模型可见提示词中没有 HTML 注释、隐藏标记，也没有会被解析回状态的标题。
 
 协调时，插件扫描带工作区来源的 `user/message` 事件，并派生每个可见作用域的最新状态。一个简短的逐会话待处理映射只会在不可变的顶层 `tools/result` 证明某个 `additionalContexts` 条目经过所有 post-execute 监听器后仍然保留时开始记录；随后，它覆盖循环将该上下文追加到日志之前的间隔。每个条目记录开启状态的 `{ turn, step }`：如果相同的持久 `user/message` 出现在其序列边界或之后，该条目得到确认并被移除；如果匹配的 `step/end` 先到达，则说明循环丢弃了上下文缓冲区，插件会同时移除待处理条目及其版本缓存快速路径。嵌套的 Code Mode 结果会把变更暂存在父级的不透明执行 token 下，确保一次运行中的重复子分发不会产生重复项；父级结果会回滚这份临时状态，并且只提交外层策略保留的上下文。
 
-路径和 digest 均未变化时会被抑制。日志中的移除操作是一条墓碑记录，因此重新出现的候选项会成为新的 `set`。恢复操作从持久化元数据继续工作。如果压缩从可见表面移除某条指令事件，该状态不再抑制后续加载，这与模型已经无法看见它的事实一致。只有真正纳入字节预算的变更才会进入元数据或待处理状态，因此被省略的文件在之后的触碰中仍有资格加载。
+路径和 digest 均未变化时会被抑制。日志中的移除操作是一条墓碑记录，因此重新出现的候选项会成为新的 `set`。恢复操作从持久化元数据继续工作：兼容的可见基线会提供比较状态，而不会导致再次追加完整基线。如果压缩从可见表面移除某条指令事件，该状态不再抑制后续加载，这与模型已经无法看见它的事实一致。只有真正纳入字节预算的变更才会进入元数据或待处理状态，因此被省略的文件在之后的触碰中仍有资格加载。
 
-只有当初始基线事件仍在可见会话表层中时，其类型化变更才用作比较状态。后续成功的文件系统触碰会在压缩后重新添加未变化的基线 scope，或把基线编辑或移除操作追加为动态消息；它绝不重写原始事件。内存中的 scope 标记和提供方版本 cache 只用于选择探测对象并加速探测，因此二者都不能抑制模型已无法看见的上下文。恢复时准备基线的过程中，插件还会协调可见的动态作用域，因此 agent 离线期间发生的嵌套变更可以在第一次恢复请求前追加更新。
+只有当初始基线事件仍在可见会话表层中时，其类型化变更才用作比较状态。恢复会保留兼容的基线，并对账当前基线和可见的动态 scope，因此 agent 离线期间的变更会在第一次恢复请求前追加为转换。当压缩遮蔽基线事件时，下一次进入步骤的 pre-step 会组合完整的当前基线，并在同一请求中记录它；也可以改由一次成功的文件系统触碰重新添加未变化的基线 scope，或把基线编辑或移除操作追加为动态消息。两条路径都不会重写原始事件。内存中的 scope 标记和提供方版本 cache 只用于选择探测对象并加速探测，因此二者都不能抑制模型已无法看见的上下文。
 
-系统刻意不使用文件监视器。检测发生在下一次成功的结构化文件系统触碰或恢复时的基线准备。提供方失败不会产生移除；只有该作用域中的全部已配置候选项都成功完成探测后，系统才接受「不存在」这一结论。
+系统刻意不使用文件监视器。检测发生在下一次成功的结构化文件系统触碰、恢复时的基线对账，或进入步骤的 pre-step 恢复被遮蔽的基线时。提供方失败不会产生移除；只有该作用域中的全部已配置候选项都成功完成探测后，系统才接受「不存在」这一结论。
 
 ### 字节预算与有界读取
 
@@ -82,7 +82,7 @@ shell 命令不会触发发现。本地 bash 调用会启动全新的 shell，
 
 仓库文本仍是不受信任的输入。低权威 user 角色框架、显式优先级说明和分隔符转义可以降低风险，但无法消除提示词注入。跟随候选符号链接到目标，会把该攻击面扩大至树外内容；因此，把 `ctx.fs` 限制在可信根目录内的权限与沙箱层才是真正的边界，它们让系统把工作区文件当作数据而不是权威（[跟随指令符号链接记录](2026-07-21-follow-instruction-symlinks.md)负责说明残余风险）。
 
-系统由事件驱动，而不是文件监视器驱动。除非文件系统变更通过结构化工具完成，否则编辑不会在确切的文件系统变更时刻可见；外部文件变更会在下一次成功的结构化触碰或恢复时被发现。这使设计保持确定性并且与提供方无关。
+系统由事件驱动，而不是文件监视器驱动。除非文件系统变更通过结构化工具完成，否则编辑不会在确切的文件系统变更时刻可见；外部文件变更会在下一次成功的结构化触碰、恢复对账，或恢复被遮蔽的基线时被发现。这使设计保持确定性并且与提供方无关。
 
 ## 延后事项
 

.agents/notes/implemented/feature/2026-06-29-todo-write-tool.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/2026-06-29-todo-write-tool.md
-2026-06-29-todo-write-tool.md: 288932f641a37c13ea6beeb069ac360c4a8447c1
-2026-06-29-todo-write-tool.zh.md: 2eced2e1670c20726989137d441d1c78df289641
+2026-06-29-todo-write-tool.md: c4cf64b7876bd8b80df80fe6fc27005715f96b5e
+2026-06-29-todo-write-tool.zh.md: 1f797c5517df8affd66addecf2a8db43c1da3332

.agents/notes/implemented/feature/2026-06-29-todo-write-tool.md

@@ -6,7 +6,7 @@ English | [中文](2026-06-29-todo-write-tool.zh.md)
 
 ## Problem
 
-The harness gives the model bash and subagent tools but no way to record a structured task list. A todo list serves two co-equal purposes: it steers the model to plan multi-step work and keep the active task unambiguous (at most one active, exactly one while work remains), and it gives an interactive host a live progress checklist. Every reference coding agent surveyed (claude-code, opencode, codex, oh-my-pi, pi) ships some form of this; the harness had nothing.
+The harness gives the model bash and subagent tools but no way to record a structured task list. A todo list serves two co-equal purposes: it steers the model to plan multi-step work and keep the active work unambiguous, and it gives an interactive host a live progress checklist. Every reference coding agent surveyed (claude-code, opencode, codex, oh-my-pi, pi) ships some form of this; the harness had nothing.
 
 ## Decision
 
@@ -34,7 +34,7 @@ Each list belongs to the calling agent session, and non-agent calls are rejected
 
 ### Validation: the cheap middle
 
-The schema enforces type/required/enum. Beyond that, `execute` rejects empty or duplicate `content` and more than one `in_progress` task. claude-code leaves single-in-progress to the prompt; oh-my-pi enforces it in code. We take the middle: enforce the cheap invariants that make a plan *coherent* (no blank tasks, no dupes, at most one active), but leave ordering and the discipline of keeping the list current to the model via the tool description. A rejected write returns an `isError` result so the model self-corrects.
+The schema enforces type/required/enum. Beyond that, `execute` rejects empty or duplicate `content` and, when `allowParallelInProgress` is `false`, more than one active task. Ordering and keeping the list current remain model disciplines expressed in the tool description. A rejected write returns an `isError` result so the model self-corrects. The required deployment policy and the durable invariant's independence from it are owned by the [parallel in-progress Agent Note](2026-07-26-todo-parallel-in-progress.md).
 
 ## Why no cordis-catalog entry / no `@mode`
 

.agents/notes/implemented/feature/2026-06-29-todo-write-tool.zh.md

@@ -6,7 +6,7 @@ Status: implemented
 
 ## 问题
 
-harness 为模型提供了 bash 和 subagent 工具，却没有办法记录结构化的任务列表。todo 列表有两个同等重要的用途：引导模型规划多步骤工作并保持当前活跃任务明确（最多一个活跃，有剩余工作时恰好一个）；同时为交互式宿主提供实时进度清单。调研的所有参考编码 agent（智能体），包括 claude-code、opencode、codex、oh-my-pi 和 pi，都提供了某种形式的此功能；本 harness 此前没有。
+harness 为模型提供了 bash 和 subagent 工具，却没有办法记录结构化的任务列表。todo 列表有两个同等重要的用途：引导模型规划多步骤工作并保持当前活跃任务明确；同时为交互式宿主提供实时进度清单。调研的所有参考编码 agent（智能体），包括 claude-code、opencode、codex、oh-my-pi 和 pi，都提供了某种形式的此功能；本 harness 此前没有。
 
 ## 决策
 
@@ -34,7 +34,7 @@ claude-code V1 的条目是 `{ content, status, activeForm }`；后来（V2）
 
 ### 校验：低成本的中间路线
 
-schema 强制 type/required/enum。在此之上，`execute` 拒绝为空或重复的 `content`，以及超过一个 `in_progress` 任务。claude-code 将单一 in_progress 交给提示词约束；oh-my-pi 在代码中强制。我们取中间路线：强制执行使计划*连贯*的低成本不变式（无空任务、无重复、最多一个活跃），但将排序和保持列表最新的纪律通过工具描述交给模型。被拒绝的写入返回 `isError` 结果，使模型自行修正。
+schema 强制 type/required/enum。在此之上，`execute` 拒绝为空或重复的 `content`，并在 `allowParallelInProgress` 为 `false` 时拒绝超过一个活跃任务。排序和保持列表最新仍通过工具描述交给模型。被拒绝的写入返回 `isError` 结果，使模型自行修正。必填的部署策略及持久日志不变式独立于该策略这一约定，由[并行 in-progress Agent Note](2026-07-26-todo-parallel-in-progress.md) 负责。
 
 ## 为何没有 cordis-catalog 条目 / 没有 `@mode`
 

.agents/notes/implemented/feature/2026-07-20-dsh-cli-personal-config.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/2026-07-20-dsh-cli-personal-config.md
-2026-07-20-dsh-cli-personal-config.md: 1fa8cda2b34b58cc7a28b722872520b68a9b7009
-2026-07-20-dsh-cli-personal-config.zh.md: e70b8914cf005e0a2e54ba2b29d3b7def84b00db
+2026-07-20-dsh-cli-personal-config.md: 10f16a1cbabdd8cd383c59ad8e09787c02d0109a
+2026-07-20-dsh-cli-personal-config.zh.md: 22435efbec8ea661c546ffd0c1aa9bb0ff2ebbb2

.agents/notes/implemented/feature/2026-07-20-dsh-cli-personal-config.md

@@ -10,6 +10,8 @@ A developer's own preferences — which provider and model the TUI uses, persona
 
 ## Decision
 
+The entry modes and the personal file's name and location below are superseded by the [profile plugin bundles decision](../architecture/2026-08-05-profile-plugin-bundles.md): `dsh` boots profiles, and the personal layer became the per-profile and home-level `cordis.patch.yml`. What survives unchanged is this note's substance — the Harness home as the machine-level layer's root, patch semantics over a shipped composition, and fail-loud parsing.
+
 Two coupled pieces, aligned with the `apps/` assembly tier proposed by the `dsh web` PR (#443):
 
 **The `dsh` CLI (`apps/cli`, npm name `@deepseek-ai/dsh`).** `apps/*` is the product-assembly tier over `packages/*` libraries. One bin dispatches the default interactive TUI, `-p`/`--prompt` headless turns, and the `web` surface. The TUI boots `examples/tui-agent/cordis.yml` (or `--config`) with the invoking directory as the workspace. The committed `bin/dsh` launcher resolves the checkout through its own real path and runs the app with tsx's ESM hook; the [source-launch decision](../architecture/2026-07-29-dsh-source-launch-tsx-esm.md) owns that contract. `pnpm run demo:tui` runs the same entry.
@@ -46,4 +48,4 @@ The TUI and Web register the exact personal path through Cordis HMR after boot.
 
 ## Testing
 
-`packages/ui/app-boot/tests/personal-config.spec.ts` pins parsing, startup application, exact-path add/failure/recovery/removal, last-good rollback, failure broadcast, and preservation of app-owned patches. `examples/tui-agent/tests/tui-keyless-smoke.e2e.ts` boots the real dsh bin with no overlay, a personal environment and UI patch, a config-only cached repository skill, and invalid personal YAML. Test launchers isolate `$DSH_HOME`, so a developer's real overlay cannot leak into fixtures.
+`packages/ui/app-boot/tests/user-patches.spec.ts` pins parsing, startup application, exact-path add/failure/recovery/removal, last-good rollback, failure broadcast, and preservation of app-owned patches. `apps/cli/tests/built-bin.e2e.ts` boots the real dsh bin over a profile and exercises the live patch layer end to end. Test launchers isolate `$DSH_HOME`, so a developer's real overlay cannot leak into fixtures.

.agents/notes/implemented/feature/2026-07-20-dsh-cli-personal-config.zh.md

@@ -10,6 +10,8 @@ Status: implemented
 
 ## Decision
 
+下文的各入口模式，以及个人文件的名称与位置，已被 [profile 插件组合包决策](../architecture/2026-08-05-profile-plugin-bundles.md)取代：`dsh` 启动 profile，个人层变成逐 profile 与 home 级的 `cordis.patch.yml`。保留不变的是本笔记的实质：以 Harness home 作为机器级层的根目录、在随附组合之上使用 patch 语义，以及解析时的大声失败。
+
 两个耦合的部分，与 `dsh web` PR（#443）提出的 `apps/` 装配层对齐：
 
 **`dsh` CLI（`apps/cli`，npm 名 `@deepseek-ai/dsh`）。** `apps/*` 是位于 `packages/*` 库之上的产品组装层。一个 bin 负责分发默认交互式 TUI、`-p`/`--prompt` 无头轮次和 `web` 界面。TUI 以调用目录为 workspace，启动 `examples/tui-agent/cordis.yml`（或 `--config` 指定的配置）。已提交的 `bin/dsh` 启动器通过自身真实路径解析 checkout，并使用 tsx 的 ESM hook 运行应用；该契约由[源码启动决策](../architecture/2026-07-29-dsh-source-launch-tsx-esm.md)维护。`pnpm run demo:tui` 运行同一入口。
@@ -46,4 +48,4 @@ TUI 和 Web 启动后通过 Cordis HMR（热模块替换）注册确切的个人
 
 ## Testing
 
-`packages/ui/app-boot/tests/personal-config.spec.ts` 固定解析、启动时应用、确切路径的新增／失败／恢复／移除、最后可用状态回滚、失败广播以及应用自有 patch 的保留。`examples/tui-agent/tests/tui-keyless-smoke.e2e.ts` 启动真实 dsh bin，覆盖无 overlay、个人环境与 UI patch、纯配置的缓存 repository skill，以及无效个人 YAML。测试启动器会隔离 `$DSH_HOME`，因此开发者的真实 overlay 不会泄漏进 fixture。
+`packages/ui/app-boot/tests/user-patches.spec.ts` 固定解析、启动时应用、确切路径的新增／失败／恢复／移除、最后可用状态回滚、失败广播以及应用自有 patch 的保留。`apps/cli/tests/built-bin.e2e.ts` 启动真实 dsh bin 并基于 profile 端到端验证实时 patch 层。测试启动器会隔离 `$DSH_HOME`，因此开发者的真实 overlay 不会泄漏进 fixture。

.agents/notes/implemented/feature/2026-07-22-durable-subagent-catalog-and-list-agents.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/2026-07-22-durable-subagent-catalog-and-list-agents.md
-2026-07-22-durable-subagent-catalog-and-list-agents.md: 0dd7eebac74689004014248c7178dba540ef4662
-2026-07-22-durable-subagent-catalog-and-list-agents.zh.md: 33b0296cf9914d1975fb1dc84564b498a09bd511
+2026-07-22-durable-subagent-catalog-and-list-agents.md: b96d6e1dd36c58af67c8e93e62515672790ad009
+2026-07-22-durable-subagent-catalog-and-list-agents.zh.md: 856bac615db84bfe2898ec0838094c6bc29f77b2

.agents/notes/implemented/feature/2026-07-22-durable-subagent-catalog-and-list-agents.md

@@ -12,6 +12,8 @@ Enumeration must cross-check immutable session lineage, descriptor validity, and
 
 ## Decision
 
+**Superseded read path.** [Subagent list identity via the projection unit](../architecture/2026-08-06-subagent-list-identity-projection.md) replaces this note's enumeration and per-child read design: `listChildren` now merges the live session store with optional session persistence directly and serves each child's mode/label from the registered `subagent` projection unit — no session-query dependency, no list-time descriptor scan — and that note owns the current listing semantics, including the diagnostic mapping. This note remains the authority for descriptor persistence, the mode-discriminated descriptor as durable identity, direct-parent authorization, and the model-facing `list_agents` projection; the trace-based read mechanics below are decision context, not current behavior.
+
 Parent-to-child enumeration is a service capability with consumer-specific projections. `SubagentService.listChildren(parentSessionId: SessionId)` ([subagent/src/index.ts](../../../../packages/subagent/subagent/src/index.ts)) does the following:
 
 - use `ctx.sessionQuery.traceSession(parentSessionId)` to obtain the parent's direct live-preferred child sessions;
@@ -33,7 +35,7 @@ Session lineage is broader than subagent identity: an ordinary `ctx.sessions.for
 
 The published logical record is also the activity source: `SessionRecord.live` means `running`, while `live: false, persisted: true` means `inactive`. Activity comes directly from the trace and causes no additional child-log load. `inactive` encodes neither successful completion nor resumability: it may describe settled one-shot history or a continuable child for which `send_message` can materialize another Activation. Conversely, `running` says only that the session is live: a live continuable Agent outside the continuation manager's matching Activation still appears as `running`, but `send_message` rejects it as an ownership conflict. A child is not visible before its session is published, and no process-local Activation entry is added as a second candidate or activity source. Listing is a snapshot that may race publication, disposal, or a later message; `send_message` remains the authoritative delivery-time operation.
 
-The subagent service keeps `sessionQuery` optional so start and follow-up remain available without it. Its public `listChildren(parentSessionId: SessionId)` method resolves the optional service and dynamically loads the optional session-query runtime only when called; ordinary subagent imports, start, and follow-up therefore do not evaluate that package. Listing belongs directly to `SubagentService`: it interprets the query's lineage, events, and live state without resolving the Activation-based continuation manager or consulting Agent registrations, Activations, or providers, so a deployment with sessions, `subagents`, and `sessionQuery` can list even when `agents` is absent. The method throws `SubagentError` with stable code `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` before loading the runtime or doing query work when the query service is absent. `@deepseek-ai/dsh-tool-subagent-control` exports separately loadable tool plugins: the `send_message` adapter requires only `subagents`, while the `list_agents` adapter requires both `subagents` and `sessionQuery` at load. A deployment may therefore use `send_message` without installing or loading session query; the list-tool fiber remains inactive until the required service is available, while another direct service consumer receives the same explicit call-time contract.
+The subagent service keeps `sessionQuery` optional so start and follow-up remain available without it. Its public `listChildren(parentSessionId: SessionId)` method resolves the optional service and dynamically loads the optional session-query runtime only when called; ordinary subagent imports, start, and follow-up therefore do not evaluate that package. Listing belongs directly to `SubagentService`: it interprets the query's lineage, events, and live state without resolving the Activation-based continuation manager or consulting Agent registrations, Activations, or providers, so a deployment with sessions, `subagents`, and `sessionQuery` can list even when `agents` is absent. The method throws `SubagentError` with stable code `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` before loading the runtime or doing query work when the query service is absent. `@deepseek-ai/dsh-tool-subagent-control` exports separately loadable tool plugins: the `send_message` adapter requires only `subagents`, while the `list_agents` adapter requires both `subagents` and `sessionQuery` at load. A deployment may therefore use `send_message` without installing or loading session query; the list-tool fiber remains inactive until the required service is available, while another direct service consumer receives the same explicit call-time contract. This dependency posture — the optional `sessionQuery`, its error code, and the list tool's load requirement — is part of the superseded read path: the current codes (`SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`, `SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE`) and the narrowed load requirement live in the [superseding note](../architecture/2026-08-06-subagent-list-identity-projection.md).
 
 `listChildren(parentSessionId, signal?)` forwards the caller's signal to `traceSession()` and the conditional exact `readEvent()` operation. `listEvents()` has no cancellation parameter, so the listing path checks the signal before and after that await and after each candidate settles. If any query operation rejects after the signal aborts, the service normalizes the result to `SubagentError` with stable code `CANCELLED`; a backend abort error or a diagnostic-mapped query error cannot escape or become a successful partial listing.
 
@@ -91,9 +93,10 @@ The first version has no child deletion operation. If later product behavior del
 ## Testing
 
 - `packages/subagent/subagent/tests/service.spec.ts` pins descriptor v2 parsing for both modes and proves an unlabeled raw start resolves a one-shot descriptor before provider dispatch. `packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts` proves the local driver appends that descriptor inside the initial turn, returns the published id when cancellation lands in the factory-to-run handoff, and keeps result and handle-disposal failures on separate channels. Delegation-tool tests pin propagation of their existing display description and preserve independent result and disposal diagnostics.
-- `packages/subagent/subagent/tests/list-children.spec.ts` pins a query-only composition with sessions, `subagents`, and `sessionQuery` but no `agents`, then drives the full real stack (agent loop, JSONL persistence, spawn/fork providers, the subagent service, and a concrete session-query service) keylessly: one-shot and continuable children from one real trace; a persisted (restart-shaped) parent target; `createdAt`-then-id ordering with authored ties; ordinary-fork and fork-seed ancestor-descriptor exclusion without diagnostics; live `running` vs persisted `inactive`; duplicate-descriptor, malformed-payload, invalid-surface, mismatched-header, and changed-read-target corruption diagnostics that leave healthy siblings visible; unsupported-version and per-child unavailable diagnostics; provider absence without child omission; compacted/uncompacted twins listing identically; grandchild exclusion; trace-phase failure failing the whole call while candidate-phase failures isolate to one child; configuration/window and unrecognized failures propagating as operation failures; forwarded trace/exact-read cancellation with stable `CANCELLED` normalization; and the `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` no-service contract. `packages/subagent/subagent/tests/optional-session-query.spec.ts` rejects eager evaluation of the optional runtime while importing the ordinary subagent surface.
-- `packages/subagent/tool-subagent-control/tests/list-agents.spec.ts` pins the `list_agents` schema (no parameters), the continuable-only projection that omits a healthy one-shot sibling while preserving diagnostics, the fixed child/diagnostic/empty text forms, an end-to-end settled-child listing with its durable label, forwarding of the tool cancellation signal, the no-agent rejection, load-time `sessionQuery` injection, and HMR disposal.
-- The keyless ACP snapshot scenario `subagent-list-agents` (examples/acp-agent) fences its second parent turn on a snapshot-only `subagent/end` marker, then executes `list_agents` for real against the subagent service, session query, and JSONL persistence, rendering `<id> [complete] — <label>`.
+- `packages/subagent/subagent/tests/list-children.spec.ts` pins the current read path against a real composition of the session store, JSONL persistence, spawn/fork providers, the subagent service, and the projection registry — no query service — keylessly: live-only listing without persistence; loud `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE` and `SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE` even with zero children; the three-rung ladder (a live child never inspected, a cold child inspected exactly once, and the cache-hit, absent-key, absent-service, and poisoned-row second-rung cases); last-wins over multiple descriptors; malformed payloads and unknown versions diagnosed as `corrupt`; a failed cold inspection as one `unavailable` diagnostic retried on the next listing; a fork seed's ancestor descriptor listed under that identity; foreign-unit fold failures contained per child as `corrupt` on both the live and cold paths; `createdAt`-then-id ordering without ordinary forks; provider absence without child omission; compacted/uncompacted twins listing identically; a persisted-listing failure failing the whole enumeration; cancellation normalized to stable `CANCELLED`; and typed stable error codes. A companion spec (retired together with the query-backed read path) rejected eager evaluation of the optional session-query runtime while importing the ordinary subagent surface.
+- `packages/subagent/tool-subagent-control/tests/list-agents.spec.ts` pins the `list_agents` schema (no parameters), the continuable-only projection that omits a healthy one-shot sibling while preserving diagnostics, the fixed child/diagnostic/empty text forms, an end-to-end settled-child listing with its durable label, forwarding of the tool cancellation signal, the no-agent rejection, the narrowed load requirement without `sessionQuery`, and HMR disposal.
+- The keyless ACP snapshot scenario `subagent-list-agents` (examples/acp-agent) fences its second parent turn on a snapshot-only `subagent/end` marker, then executes `list_agents` for real against the subagent service, the projection registry, and JSONL persistence, rendering `<id> [complete] — <label>`.
+- The keyless snapshot scenario `subagent-diagnostic` (examples/headless-agent) pins the current listing's model-visible diagnostic classification, including a descriptor-less settled child surfacing as a `corrupt` diagnostic.
 - The keyless ACP snapshot scenario `subagent-published-run-failure` publishes a real one-shot child, injects independent run-result and handle-disposal failures, and preserves both diagnostics in the parent tool result.
 
 ## Consequences

.agents/notes/implemented/feature/2026-07-22-durable-subagent-catalog-and-list-agents.zh.md

@@ -12,6 +12,8 @@ Status: implemented
 
 ## 决策
 
+**列表读路径已被取代。**[subagent 列表经投影单元读取身份](../architecture/2026-08-06-subagent-list-identity-projection.md)取代了本记录的枚举与逐 child 读取设计:`listChildren` 现在直接合并存活会话存储与可选的会话持久化,并从注册的 `subagent` projection unit 读取每个 child 的 mode/label——不依赖会话查询,也不在列表时扫描描述符;当前的列表语义(含 diagnostic 映射)以该记录为准。本记录仍是描述符持久化、以 mode 判别的描述符作为持久身份、直接 parent 鉴权与面向模型的 `list_agents` 投影的权威;下文基于追踪的读取机制是决策背景,不再是当前行为。
+
 parent 到 child 的枚举是一项带消费方专用投影的服务功能。`SubagentService.listChildren(parentSessionId: SessionId)`（[subagent/src/index.ts](../../../../packages/subagent/subagent/src/index.ts)）执行以下操作：
 
 - 使用 `ctx.sessionQuery.traceSession(parentSessionId)` 获取 parent 的直接且实时优先的 child 会话；
@@ -33,7 +35,7 @@ parent 到 child 的枚举是一项带消费方专用投影的服务功能。`Su
 
 已发布的逻辑记录同时也是活动状态来源：`SessionRecord.live` 表示 `running`，而 `live: false, persisted: true` 表示 `inactive`。活动状态直接来自追踪结果，不会导致额外加载 child 日志。`inactive` 既不表示执行成功，也不表示可恢复：它可能表示已结算的一次性历史，也可能表示 `send_message` 可以为其物化另一次 Activation 的可继续 child。反过来，`running` 只表示会话存活：位于继续执行管理器对应 Activation 之外的存活可继续 Agent 仍会显示为 `running`，但 `send_message` 会将其作为所有权冲突拒绝。child 会话发布前不可见，也不会添加进程内 Activation 条目作为第二个候选来源或活动状态来源。列表查询是一份快照，可能与发布、dispose 或后续消息发生竞态；`send_message` 仍是消息送达时的权威操作。
 
-subagent 服务将 `sessionQuery` 保持为可选依赖，因此没有该服务时仍可执行 start 和 follow-up。其公开的 `listChildren(parentSessionId: SessionId)` 方法只在被调用时才会解析这个可选服务，并动态加载可选的会话查询运行时；因此，普通 subagent 导入、start 和 follow-up 都不会触发该包求值。列表查询直接由 `SubagentService` 负责：它解释查询返回的谱系、事件和存活状态，无需解析基于 Activation 的继续执行管理器，也不会查询 Agent 注册信息、Activation 或提供方；因此，仅包含会话、`subagents` 和 `sessionQuery` 的部署即使缺少 `agents` 也能执行列表查询。如果查询服务缺失，该方法会在加载运行时或执行查询工作前抛出 `SubagentError`，并携带稳定错误码 `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE`。`@deepseek-ai/dsh-tool-subagent-control` 导出可分别加载的工具插件：`send_message` 适配器只要求 `subagents`，而 `list_agents` 适配器在加载时同时要求 `subagents` 和 `sessionQuery`。因此，部署可以在既不安装也不加载会话查询的情况下使用 `send_message`；列表工具 fiber 会在必需服务可用前保持未激活状态，而其他直接服务消费方会收到同一项明确的调用时契约。
+subagent 服务将 `sessionQuery` 保持为可选依赖，因此没有该服务时仍可执行 start 和 follow-up。其公开的 `listChildren(parentSessionId: SessionId)` 方法只在被调用时才会解析这个可选服务，并动态加载可选的会话查询运行时；因此，普通 subagent 导入、start 和 follow-up 都不会触发该包求值。列表查询直接由 `SubagentService` 负责：它解释查询返回的谱系、事件和存活状态，无需解析基于 Activation 的继续执行管理器，也不会查询 Agent 注册信息、Activation 或提供方；因此，仅包含会话、`subagents` 和 `sessionQuery` 的部署即使缺少 `agents` 也能执行列表查询。如果查询服务缺失，该方法会在加载运行时或执行查询工作前抛出 `SubagentError`，并携带稳定错误码 `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE`。`@deepseek-ai/dsh-tool-subagent-control` 导出可分别加载的工具插件：`send_message` 适配器只要求 `subagents`，而 `list_agents` 适配器在加载时同时要求 `subagents` 和 `sessionQuery`。因此，部署可以在既不安装也不加载会话查询的情况下使用 `send_message`；列表工具 fiber 会在必需服务可用前保持未激活状态，而其他直接服务消费方会收到同一项明确的调用时契约。这一段的依赖姿态——可选 `sessionQuery`、其错误码与列表工具的加载要求——同属被取代的读路径：现行错误码（`SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE`、`SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE`）与收窄后的加载要求以[取代记录](../architecture/2026-08-06-subagent-list-identity-projection.md)为准。
 
 `listChildren(parentSessionId, signal?)` 会把调用方的取消信号转发给 `traceSession()` 和条件性精确 `readEvent()` 操作。`listEvents()` 不接受取消参数，因此列表查询路径会在等待该操作的前后，以及每个候选处理完成后检查信号。如果取消信号触发后有查询操作以拒绝结算，服务会将结果归一化为 `SubagentError`，并携带稳定错误码 `CANCELLED`；后端中止错误或可映射为 diagnostic 的查询错误均不会逃逸，也不会使调用以成功的部分列表返回。
 
@@ -91,9 +93,10 @@ diagnostic 是瞬时查询结果，不属于会话事件或目录状态。推导
 ## 测试
 
 - `packages/subagent/subagent/tests/service.spec.ts` 固定两种模式下的描述符 v2 解析，并证明无标签的底层启动会在分发给提供方之前解析出一次性描述符。`packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts` 证明本地驱动会在初始轮次内追加该描述符，在取消落入工厂到 run 的交接窗口时返回已发布 id，并让结果与句柄释放失败保留在独立通道中。委派工具测试固定其现有显示说明的传递，并保留相互独立的结果与 dispose diagnostic。
-- `packages/subagent/subagent/tests/list-children.spec.ts` 先固定一个只有会话、`subagents` 和 `sessionQuery` 而没有 `agents` 的纯查询组合，再以无密钥方式驱动完整真实栈（agent loop、JSONL 持久化、spawn/fork 提供方、subagent 服务，以及一个具体的会话查询服务）：来自同一真实追踪的一次性与可继续 child；只存在于持久化存储中（重启形态）的 parent 目标；带有人工构造并列项的按 `createdAt` 再按 id 排序；排除普通 fork 和 fork seed 中祖先描述符且不产生 diagnostic；存活 `running` 与持久化 `inactive` 的对比；重复描述符、载荷格式错误、无效 surface、header 不匹配和读取目标已变化的损坏 diagnostic 均不隐藏健康的 sibling；不受支持版本与逐 child unavailable diagnostic；提供方缺失时不排除 child；压缩与未压缩的孪生 child 列表结果一致；排除孙代会话；追踪阶段失败导致整次调用失败而候选阶段失败只隔离到单个 child；配置／窗口错误和无法识别的失败作为操作失败向上传播；转发 trace／精确读取取消并稳定归一化为 `CANCELLED`；以及 `SUBAGENT_CONTROL_SESSION_QUERY_UNAVAILABLE` 缺服务契约。`packages/subagent/subagent/tests/optional-session-query.spec.ts` 会在导入普通 subagent surface 时拒绝对可选运行时的 eager 求值。
-- `packages/subagent/tool-subagent-control/tests/list-agents.spec.ts` 固定 `list_agents` 的 schema（无参数）、只保留可继续 child 且排除健康的一次性 sibling、同时保留 diagnostic 的投影、child／diagnostic／空结果的固定文本形式、带持久化 label 的已结束 child 端到端列表、工具取消信号的转发、无调用 agent 时的拒绝、加载时的 `sessionQuery` 注入，以及 HMR dispose。
-- 无密钥 ACP 快照场景 `subagent-list-agents`（examples/acp-agent）使用仅限快照的 `subagent/end` 标记为第二个 parent 轮次设置边界，随后针对 subagent 服务、会话查询和 JSONL 持久化真实执行 `list_agents`，渲染 `<id> [complete] — <label>`。
+- `packages/subagent/subagent/tests/list-children.spec.ts` 针对由会话存储、JSONL 持久化、spawn/fork 提供方、subagent 服务与投影注册表构成的真实组合——不含查询服务——以无密钥方式钉住现行读取路径：无持久化时的仅存活列表；零 children 也响亮报 `SUBAGENT_CONTROL_PROJECTIONS_UNAVAILABLE` 与 `SUBAGENT_CONTROL_SESSION_STORE_UNAVAILABLE`；三级阶梯（存活 child 从不检查、冷 child 恰好检查一次，以及缓存命中、key 缺席、服务缺席、行中毒四个第二级用例）；多描述符 last-wins 取末者；载荷格式错误与未知版本诊断为 `corrupt`；冷检查失败成一条 `unavailable` diagnostic 并在下次列表重试；fork seed 中的祖先描述符按该身份列出；外部 unit 折叠失败在存活与冷两条路径上按 child 收纳为 `corrupt`；按 `createdAt` 再按 id 排序且不列普通 fork；提供方缺失时不排除 child；压缩与未压缩的孪生 child 列表结果一致；持久化列表失败使整次枚举失败；取消稳定归一化为 `CANCELLED`；以及带类型的稳定错误码。一个伴随规格(已随查询式读取路径一起退役)曾在导入普通 subagent surface 时拒绝对可选 session-query 运行时的 eager 求值。
+- `packages/subagent/tool-subagent-control/tests/list-agents.spec.ts` 固定 `list_agents` 的 schema（无参数）、只保留可继续 child 且排除健康的一次性 sibling、同时保留 diagnostic 的投影、child／diagnostic／空结果的固定文本形式、带持久化 label 的已结束 child 端到端列表、工具取消信号的转发、无调用 agent 时的拒绝、收窄后的加载要求（不再注入 `sessionQuery`），以及 HMR dispose。
+- 无密钥 ACP 快照场景 `subagent-list-agents`（examples/acp-agent）使用仅限快照的 `subagent/end` 标记为第二个 parent 轮次设置边界，随后针对 subagent 服务、投影注册表和 JSONL 持久化真实执行 `list_agents`，渲染 `<id> [complete] — <label>`。
+- 无密钥快照场景 `subagent-diagnostic`（examples/headless-agent）钉住现行列表的模型可见诊断分类，包括无描述符的定局 child 以 `corrupt` diagnostic 出现。
 - 无密钥 ACP 快照场景 `subagent-published-run-failure` 会发布一个真实的一次性 child，注入相互独立的 run result 与 handle dispose 失败，并在 parent 工具结果中保留两项 diagnostic。
 
 ## 影响

.agents/notes/implemented/feature/2026-07-23-web-todo-display.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/2026-07-23-web-todo-display.md
-2026-07-23-web-todo-display.md: d5939c0fa6ab83a61f1932622c4b6a150ccaded7
-2026-07-23-web-todo-display.zh.md: b1ecf679efe81016e00c5c1221c04730b3ce98bb
+2026-07-23-web-todo-display.md: 9e6e4914cd24d1db9271baa3d3fb6fdc56a9ac65
+2026-07-23-web-todo-display.zh.md: 5a5ac554b37c255a7d8c1ce821ebeb1b3f9f091f

.agents/notes/implemented/feature/2026-07-23-web-todo-display.md

@@ -18,11 +18,11 @@ Consume `todo/write` as a Session side effect, not a surface node, and render it
 
 ### TodoPanel: the durable list as a persistent strip
 
-The panel mounts through the `conversation.input.dock` slot (a plain registrant plugin, `todoDockEntry`, using `ctx.slots.inject` with no `ConversationService` edge, `order: -1` above the queue rows), hidden while empty, collapsible to a header of title + `"<done>/<total> tasks · <n> in progress"` (no in-progress content hint when collapsed). Status glyphs are the figma todo set (green check ring / blue fading ring / dashed pending ring) on a tip-surface card (`--dsw-specific-tip`, 14px radius, `width: calc(100% - 88px)` / `max-width: 776px` centered; InputBar top pad 6px is the gap to the composer card). It reads `snapshot.todos` via the standard-kit `useSession` hook the dock entry receives — no store, no service, no ctx. The inner component stays props-complete and framework-free; the dock adapter is a one-line wrapper.
+The panel mounts through the `conversation.input.dock` slot (a plain registrant plugin, `todoDockEntry`, using `ctx.slots.inject` with no `ConversationService` edge, `order: 0` above the queue rows), hidden while empty, collapsible to a header of title + `·`-joined per-status counts (localized, `1 completed · 2 in progress · 1 pending`, zero-count segments omitted; no in-progress content hint when collapsed). Status glyphs are the figma todo set (green check ring / blue fading ring / dashed pending ring) on a tip-surface card (`--dsw-specific-tip`, 14px radius, `width: calc(100% - 88px)` / `max-width: 776px` centered; InputBar top pad 6px is the gap to the composer card). It reads the host-computed `todos` projection via the standard-kit `useProjection` hook the dock entry receives — no store, no service, no ctx. The inner component stays props-complete and framework-free; the dock adapter is a one-line wrapper.
 
 ### TodoRow: the per-call row through the keyed toolview slot
 
-The dedicated `todo_write` chat row is a plain registrant plugin (`todoToolview`, mounted from `apply`) that registers into the keyed `conversation.chat.toolview` slot through `ctx.slots.inject`, the same declaration-lifetime posture as the bash sample but a product registration. The summary derives from call args (`N/M done · active item`); unparseable args fall back to the generic row summary; clicking opens the details column with the raw args. No `ToolEventView` is added for todo — presentation is client-owned, and the durable list renders from the session event, not the tool card.
+The dedicated `todo_write` chat row is a plain registrant plugin (`todoToolview`, mounted from `apply`) that registers into the keyed `conversation.chat.toolview` slot through `ctx.slots.inject`, the same declaration-lifetime posture as the bash sample but a product registration. The summary derives from call args (`N/M done · first active item`, with a `+<n>` count of the other active ones in `ToolRow`'s non-shrinking `summarySuffix` slot); unparseable args fall back to the generic row summary; clicking opens the details column with the raw args. No `ToolEventView` is added for todo — presentation is client-owned, and the durable list renders from the session event, not the tool card.
 
 ## Alternatives considered
 
@@ -33,4 +33,4 @@ The dedicated `todo_write` chat row is a plain registrant plugin (`todoToolview`
 
 ## Consequences
 
-Replay correctness is owned by one code path: any future change to window rebuild keeps todos consistent for free, and the fixture (fx-alpha turn 65) plus the assembled keyless snapshot (`apps/web/tests/todo-display.snapshot.ts`) pin the full chain (row summary and state, dock panel content, collapse round-trip) over the built client graph. `todos` is a required `ConversationSnapshot` field, so scripted fakes in specs must carry it. The TUI panel shares the same turn-scoped lifetime (the automation-only ACP bridge deliberately omits todo presentation); the web surfaces render the same event, adding one wire field and no new event type. That field is how cold-load reconstruction stays host-backed: the tail history page carries `todos` — the full-log standing plan (latest `todo/write` with no later `turn/start`), computed independently of the page window (the same backscan posture the view pairing uses) — so a reopened session restores the plan when it still stands and the last write precedes the window; that value survives an older-page prepend, is overwritten by any later write, clears on a later `turn/start`, and resets to empty when a tail response carries no projection.
+Replay correctness is owned by one code path: any future change to window rebuild keeps todos consistent for free, and the fixture (fx-alpha turn 71) plus `packages/client/ui-conversation/tests/todo-panel.spec.tsx` pin the full chain (row summary and state, dock panel content, collapse round-trip). `todos` is a required `ConversationSnapshot` field, so scripted fakes in specs must carry it. The automation-only ACP bridge deliberately omits todo presentation; the web surfaces render the same event, adding one wire field and no new event type. That field is how cold-load reconstruction stays host-backed: the tail history page carries `todos` — the full-log standing plan (latest `todo/write` with no later `turn/start`), computed independently of the page window (the same backscan posture the view pairing uses) — so a reopened session restores the plan when it still stands and the last write precedes the window; that value survives an older-page prepend, is overwritten by any later write, clears on a later `turn/start`, and resets to empty when a tail response carries no projection.

.agents/notes/implemented/feature/2026-07-23-web-todo-display.zh.md

@@ -18,11 +18,11 @@ Status: implemented
 
 ### TodoPanel：持久化列表作为一条常驻横条
 
-面板经 `conversation.input.dock` slot 挂载（普通注册者插件 `todoDockEntry` 使用 `ctx.slots.inject`，不依赖 `ConversationService`，`order: -1` 排在队列条上方），空列表时隐藏，可折叠为标题加 `"<已完成>/<总数> tasks · <n> in progress"` 的表头（折叠态不再附带进行中条目正文）。状态图标为 figma todo 套件（绿色勾选环／蓝色渐隐环／虚线未开始环），卡片使用 tip 表面（`--dsw-specific-tip`、14px 圆角、`width: calc(100% - 88px)`／`max-width: 776px` 居中；InputBar 顶部 6px 内边距是到输入卡的间距）。它经 dock entry 收到的标准件 `useSession` hook 读取 `snapshot.todos`——无 store、无 service、无 ctx。内部组件保持 props 完备且框架无关；dock 适配件只是一行包装。
+面板经 `conversation.input.dock` slot 挂载（普通注册者插件 `todoDockEntry` 使用 `ctx.slots.inject`，不依赖 `ConversationService`，`order: 0` 排在队列条上方），空列表时隐藏，可折叠为标题加以 `·` 连接的各状态计数的表头（本地化，形如 `1 已完成 · 2 进行中 · 1 待处理`，计数为零的段落省略；折叠态不再附带进行中条目正文）。状态图标为 figma todo 套件（绿色勾选环／蓝色渐隐环／虚线未开始环），卡片使用 tip 表面（`--dsw-specific-tip`、14px 圆角、`width: calc(100% - 88px)`／`max-width: 776px` 居中；InputBar 顶部 6px 内边距是到输入卡的间距）。它经 dock entry 收到的标准件 `useProjection` hook 读取 host 计算的 `todos` 投影——无 store、无 service、无 ctx。内部组件保持 props 完备且框架无关；dock 适配件只是一行包装。
 
 ### TodoRow：经 keyed toolview slot 的逐调用行
 
-专用的 `todo_write` 对话行是一个普通注册者插件（`todoToolview`，由 `apply` 挂载），经 `ctx.slots.inject` 注册进 keyed 的 `conversation.chat.toolview` slot，遵循与 bash 样例相同的声明生命周期，但属产品级注册。摘要由调用 args 推导（`N/M done · active item`）；无法解析的 args 回退到通用行摘要；点击会以原始 args 打开 details 列。todo 不新增任何 `ToolEventView`——呈现归客户端所有，常驻列表从会话事件渲染，而非工具卡。
+专用的 `todo_write` 对话行是一个普通注册者插件（`todoToolview`，由 `apply` 挂载），经 `ctx.slots.inject` 注册进 keyed 的 `conversation.chat.toolview` slot，遵循与 bash 样例相同的声明生命周期，但属产品级注册。摘要由调用 args 推导（`N/M done · first active item`，其余活跃项的 `+<n>` 计数放在 `ToolRow` 的不收缩 `summarySuffix` 位里）；无法解析的 args 回退到通用行摘要；点击会以原始 args 打开 details 列。todo 不新增任何 `ToolEventView`——呈现归客户端所有，常驻列表从会话事件渲染，而非工具卡。
 
 ## 考虑过的替代方案
 
@@ -33,4 +33,4 @@ Status: implemented
 
 ## 后果
 
-回放正确性由一条代码路径掌管：未来对窗口重建的任何改动都会自然保持 todos 一致；fx-alpha 第 65 轮的 fixture（测试前置数据）加上组装后的无密钥快照（`apps/web/tests/todo-display.snapshot.ts`）在构建后的完整客户端依赖图中固定整条链（行摘要与状态、dock 面板内容、折叠往返）。`todos` 是 `ConversationSnapshot` 的必填字段，所以 spec 里脚本化的 fake 必须带上它。TUI 面板共用同一按轮次界定的生命周期（自动化专用的 ACP 桥接刻意不做 todo 呈现）；Web 各面渲染同一个事件，只新增一个协议字段，不新增事件类型。这个由 host 提供的字段正是冷加载重建的依据：history 尾页附带 `todos`——全量 log 上当前有效的计划（其后没有更晚 `turn/start` 的最近一次 `todo/write`），独立于分页窗口计算（与 view 配对同一种 backscan 姿势）——因此重开会话时若计划仍然有效且最后一次写入落在窗口之前，计划也照常恢复；该值跨往前翻页保留，之后的任何写入照常覆盖，更晚的 `turn/start` 会清空，而尾页响应不带投影时复位为空。
+回放正确性由一条代码路径掌管：未来对窗口重建的任何改动都会自然保持 todos 一致；fx-alpha 第 71 轮的 fixture（测试前置数据）加上 `packages/client/ui-conversation/tests/todo-panel.spec.tsx` 固定整条链（行摘要与状态、dock 面板内容、折叠往返）。`todos` 是 `ConversationSnapshot` 的必填字段，所以 spec 里脚本化的 fake 必须带上它。自动化专用的 ACP 桥接刻意不做 todo 呈现；Web 各面渲染同一个事件，只新增一个协议字段，不新增事件类型。这个由 host 提供的字段正是冷加载重建的依据：history 尾页附带 `todos`——全量 log 上当前有效的计划（其后没有更晚 `turn/start` 的最近一次 `todo/write`），独立于分页窗口计算（与 view 配对同一种 backscan 姿势）——因此重开会话时若计划仍然有效且最后一次写入落在窗口之前，计划也照常恢复；该值跨往前翻页保留，之后的任何写入照常覆盖，更晚的 `turn/start` 会清空，而尾页响应不带投影时复位为空。

.agents/notes/implemented/feature/2026-07-26-todo-parallel-in-progress.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/2026-07-26-todo-parallel-in-progress.md
+2026-07-26-todo-parallel-in-progress.md: 2805ef894050d1b1cffe06fce59a4984d463f8d1
+2026-07-26-todo-parallel-in-progress.zh.md: 16b32daa05b10f24eacde4cec9622b2159cdef09

.agents/notes/implemented/feature/2026-07-26-todo-parallel-in-progress.md

@@ -0,0 +1,54 @@
+# Agent Note: Allow several `in_progress` todos at once
+
+Status: implemented
+
+English | [中文](2026-07-26-todo-parallel-in-progress.zh.md)
+
+## Problem
+
+The [original `todo_write` design](2026-06-29-todo-write-tool.md) enforced at most one `in_progress` task per list, both in `execute` and in the durable-log invariant. That invariant assumes sequential work, but the harness runs genuinely parallel work — concurrent subagents through the delegation tool, background bash commands, workflow fan-out — and a list that can name only one active task cannot represent it. The model was forced to either mislabel parallel tasks as `pending` or merge them into one vague item, and the UI progress checklist under-reported what was actually running.
+
+## Decision
+
+Make the single-`in_progress` cap a deployment policy instead of a fixed rule, requiring every composition to choose:
+
+- `packages/todo/tool-todo/src/index.ts` gains the required `Config.allowParallelInProgress` field. At `true`, `execute` accepts any number of active items and the description instructs the model to mark every actively-worked task — several during parallel work, one for sequential work — keeping at least one while work remains. At `false`, the description asks for exactly one and `execute` rejects a call marking more.
+- The durable-log invariant in `packages/todo/tool-todo/src/invariant.ts` no longer rejects snapshots with several active items, and does not follow the config, so previously-persisted logs are unaffected and parallel snapshots replay cleanly under either policy.
+
+The remaining coded invariants are unchanged: non-empty trimmed unique `content`, valid status enum. This supersedes the "at most one active" clause of the [original design's validation decision](2026-06-29-todo-write-tool.md); the rest of that Agent Note (whole-list replace, log-backed state, single owner) stands.
+
+## Why guidance, not a parallelism-aware invariant
+
+A coded invariant can only see the list, not the runtime: whether two `in_progress` items are legitimate depends on whether work is actually running concurrently, which the tool cannot observe. Enforcing a cap was therefore wrong in exactly the cases parallelism made it matter, and any replacement (for example, capping active items at the live subagent count) would couple the tool to runtimes it deliberately knows nothing about. The discipline of matching `in_progress` marks to genuinely concurrent work moves to the tool description, the same place ordering and list freshness already live.
+
+## The policy is a deployment choice
+
+Whether concurrent active tasks are legitimate depends on runtime concurrency the tool cannot observe — but whether a deployment's agents ever run work concurrently is knowable at composition time. That makes the policy a required `Config` field rather than a constant or default: every cordis.yml composition sets `allowParallelInProgress` deliberately, choosing `true` for agents that may fan out work or `false` for the single-active discipline.
+
+The flag moves the model-facing instruction and the accepted input together. Splitting them would be the bug: a description asking for one active task while `execute` accepts several teaches the model a rule the tool does not hold, and the reverse rejects calls the description invited. Only the active-status clause of the description varies, because that is the only instruction the policy changes.
+
+The durable-log invariant deliberately does NOT follow the flag. A log written while parallel work was allowed must still replay after a deployment tightens the policy, so tying `invariant.ts` to the current config would reject history that was valid when it was written. The invariant stays silent on the active count; the tool is where the policy applies, at the moment of the write.
+
+## Alternatives considered
+
+- **Keep the cap and add an explicit parallel opt-in flag** — an extra argument on every call to serve the common case; the flag would be noise for sequential work and still unverifiable.
+- **Cap active items at a configured maximum** — any fixed number is arbitrary. This is why the config field is a boolean policy switch and not a count: "may several tasks be active" is a property of the deployment, while "at most N" invents a threshold nothing can justify.
+- **Hardcode the parallel policy** — the first revision of this branch did, which is what made `allowParallelInProgress` necessary: a deployment running strictly sequential agents had no way back to the discipline it wanted.
+
+## The display surfaces are part of the change
+
+Lifting the cap makes a list shape reachable that no renderer had ever received, so this branch stacks on the [web todo display](2026-07-23-web-todo-display.md) rather than landing beside it: both change `tool-todo`, and the GUI is where a parallel plan becomes visible. Two web sites derived their one-line summary with `todos.find(t => t.status === 'in_progress')` — the collapsed plan-strip header and the `todo_write` row — and under the old cap that `find` was total, since at most one item could match. With several active it silently dropped every active item but the first: a four-item plan with three running tasks collapsed to the name of one, and the row read `1/4 已完成 · <one task>` while two others were in flight. The expanded list was always correct (it maps every item), which is why neither PR's tests caught it — only the collapsed header and the row lost information. The panel redesign in [#740](https://github.com/deepseek-harness/deepseek-harness/pull/740) has since replaced the collapsed header's named hint with `·`-joined per-status counts (localized, `1 completed · 2 in progress · 1 pending`, zero-count segments omitted), which reports parallel work correctly and needs no name to truncate; the row is the one site this branch still had to fix.
+
+The row takes `planSummary` in `toolviews/plan-summary.ts`. It names the first active item and counts the rest, so the row reports how many tasks are running instead of implying one. Naming every active item was rejected: the row is a single line, and an unbounded join would overflow it — the count degrades predictably where a list does not. The derivation sits inside the toolviews domain rather than in `contract/`, the inter-domain face: the panel computes its own counts inline and shares nothing with the row, so a contract module would declare a sharing relationship that no longer exists.
+
+`planSummary` returns the name and the count as separate fields rather than one joined string, because the row truncates its summary with `overflow: hidden` / `text-overflow: ellipsis`. A count appended to the task name sits at the far end of the truncatable text, so exactly the narrow viewports and long task names that make the count informative are the ones that clip it away, leaving a parallel plan indistinguishable from a sequential one. The row therefore hands the count to the shared `ToolRow` as `summarySuffix`, a non-shrinking slot beside the ellipsized summary text; a pre-joined string could not express that split, and pushing the count in front of the name was rejected because the task name is what the reader is looking for first.
+
+`summarySuffix` is a slot on `ToolRow` rather than markup owned by the todo row: every toolview renders through that shared component, whose `summary` is a plain ellipsized string with no place for a fragment that must survive the clip. Sitting outside the `.summary` rule, the suffix repeats that rule's `font-size` and `line-height` — the web shell leaves body text at the browser default rather than the row's 14px, so an unstyled span renders visibly larger than the text beside it on a 24px row. An error row drops the suffix, because its collapsed summary is the failure line rather than anything derived from the call args.
+
+## Deferred
+
+Two review findings are recorded here rather than fixed on this branch. The `summarySuffix` span carries no accessible name, so a screen reader reads the count without its noun (`… 实现 fixture 样本 +1`); naming it introduces localized copy with its own test contract, which belongs to an accessibility pass over the whole `ToolRow` summary line rather than to one row. And when the *first* active item's content is unusable — missing, mistyped, or blank once trimmed — the row drops the active clause and the count with it, so a parallel plan renders as bare counts; skipping forward to the first usable active item was rejected because call args are an explicitly unvalidated boundary where model order is the only ordering the row can honour, and dropping the unusable clause alone keeps the `done`/`total` counts, which are trustworthy regardless.
+
+## Consequences
+
+A todo list can now faithfully mirror parallel execution, and every surface renders several active markers at once: the plan strip's header counts the active items, and the row needed the derivation above. A composition that sets `allowParallelInProgress: true` no longer rejects a formerly-invalid snapshot shape; one that sets `false` keeps the old rejection, and the durable-log invariant accepts both. The model-facing description changed, which re-recorded the tool-catalog page and every snapshot sidecar carrying the todo schema. No count is recorded here: the set grows with every pinning scenario that lands, and the two point-in-time censuses this note previously carried were both stale within days. The operative rule is that a branch changing the tool description must refresh whichever sidecars landed after it branched — including the numbered `tool-schemas.<n>.expected.json` files pinning a subagent class, whose schemas the parent scenario does not cover — and `pnpm run test:snapshot:refresh` does it keylessly over the whole corpus. The web fixture's todo sample now runs two items `in_progress`, so both fixture-driven surfaces render a parallel plan. `packages/client/ui-conversation/tests/todo-panel.spec.tsx` pins the row summary and the plan strip over src, the ACP `todo-write` scenario records a three-todo plan with two active, and `apps/web/tests/todo-row.snapshot.ts` pins both surfaces in the assembled application — booted from the built `packages/client/*/lib/client.js` bundles, so it is the one place the keyed registration and the bundled wiring are under test. That last file records `summary`, `suffix`, and the strip's header as separate fields, so folding the `+N` count back into the summary string changes the expected output even though the concatenated text would read the same.

.agents/notes/implemented/feature/2026-07-26-todo-parallel-in-progress.zh.md

@@ -0,0 +1,54 @@
+# Agent Note: 允许同时存在多个 `in_progress` todo
+
+Status: implemented
+
+[English](2026-07-26-todo-parallel-in-progress.md) | 中文
+
+## 问题
+
+[原始 `todo_write` 设计](2026-06-29-todo-write-tool.md)在 `execute` 和持久日志不变式中都强制每个列表至多一个 `in_progress` 任务。该不变式假设工作是顺序进行的，但 harness 会运行真正并行的工作（通过委派工具启动的并发 subagent、后台 bash 命令、工作流扇出），而一个只能标出单个活跃任务的列表无法表示这种情况。模型被迫要么把并行任务错误标记为 `pending`，要么把它们合并成一个含糊的条目，导致 UI 进度清单少报了实际正在运行的工作。
+
+## 决策
+
+把单一 `in_progress` 上限从固定规则改为部署策略，并要求每个组合都作出选择：
+
+- `packages/todo/tool-todo/src/index.ts` 新增必填的 `Config.allowParallelInProgress` 字段。为 `true` 时，`execute` 接受任意数量的活跃条目，描述指示模型把每个正在处理的任务标记为 `in_progress`（并行工作时可以有多个，顺序工作时只有一个），并在仍有工作未完成时至少保留一个；为 `false` 时，描述要求恰好一个，`execute` 拒绝标记更多的调用。
+- `packages/todo/tool-todo/src/invariant.ts` 中的持久日志不变式不再拒绝含多个活跃条目的快照，且不跟随该配置，因此此前持久化的日志不受影响，并行快照在任一策略下都能干净回放。
+
+其余编码的不变式保持不变：`content` 去除首尾空白后非空且唯一、status 为合法枚举值。本决定取代[原始设计的校验决策](2026-06-29-todo-write-tool.md)中「至多一个活跃」的条款；该 Agent Note 的其余部分（整列表替换、日志支撑的状态、单一所有者）依然成立。
+
+## 为何用指引而非感知并行的不变式
+
+编码的不变式只能看到列表，看不到运行时：两个 `in_progress` 条目是否合理，取决于工作是否真的在并发运行，而这一点工具无法观测。因此，恰恰在并行让上限变得重要的场景里，强制上限反而是错的；任何替代方案（例如把活跃条目数限制为在线 subagent 的数量）都会把工具耦合到它有意一无所知的运行时上。把 `in_progress` 标记与真正并发的工作对应起来这一纪律，转移到工具描述中，也就是排序与列表新鲜度已经所在的地方。
+
+## 该策略是部署层的选择
+
+并发的活跃任务是否合理，取决于工具无法观测的运行时并发情况——但一个部署的 agent 是否会并发展开工作，在组装期就是可知的。因此该策略是必填的 `Config` 字段，而非常量或默认值：每个 cordis.yml 组合都会有意设置 `allowParallelInProgress`，为可能并行展开工作的 agent 选择 `true`，或为单活跃项纪律选择 `false`。
+
+该开关会同时改变面向模型的指令与接受的输入。把两者拆开才是 bug：描述要求只保留一个活跃任务、而 `execute` 却接受多个，等于教给模型一条工具并不遵守的规则；反过来则会拒绝描述所邀请的调用。描述中只有活跃状态那一句会变化，因为这是该策略唯一改变的指令。
+
+持久日志不变式刻意**不**跟随该开关。在允许并行时写下的日志，在部署收紧策略之后仍必须可回放，因此把 `invariant.ts` 绑定到当前配置会拒绝在写入当时合法的历史。不变式对活跃数量保持沉默；策略生效之处是工具，时机是写入的那一刻。
+
+## 曾考虑的替代方案
+
+- **保留上限并增加一个显式的并行 opt-in 标志**——为服务常见场景而给每次调用增加一个额外参数；这个标志对顺序工作而言只是噪声，而且仍然无法验证。
+- **把活跃条目限制在一个可配置的上限内**——任何固定数字都是任意的。这正是该配置字段是布尔策略开关而非数量的原因：「是否允许多个任务同时活跃」是部署的属性，而「最多 N 个」凭空发明了一个无从论证的阈值。
+- **把并行策略硬编码**——本分支的第一版就是如此，这也是 `allowParallelInProgress` 之所以必要的原因：运行严格顺序 agent 的部署没有任何办法回到它想要的纪律。
+
+## 展示面是本次改动的一部分
+
+解除上限使一种此前任何渲染器都不曾收到的列表形状变得可达，因此本分支 stack（栈叠）在 [web todo 展示](2026-07-23-web-todo-display.md)之上，而不是与之并行落地：两者都改 `tool-todo`，而 GUI 正是并行计划变得可见的地方。web 有两处用 `todos.find(t => t.status === 'in_progress')` 推导单行摘要——折叠态的计划横条表头与 `todo_write` 工具行——在旧上限下这个 `find` 是完备的，因为最多只能有一个条目匹配。一旦有多个活跃项，它会静默丢掉除第一个之外的全部活跃条目：一个四条目、三个任务在跑的计划折叠后只显示其中一个的名字，工具行读作 `1/4 已完成 · <一个任务>`，而另外两个仍在进行。展开态的列表始终正确（它遍历每个条目），这也是两个 PR 的测试都没抓到它的原因——只有折叠表头与工具行丢失了信息。其后 [#740](https://github.com/deepseek-harness/deepseek-harness/pull/740) 的面板重做已把折叠表头的具名提示换成以 `·` 连接的各状态计数（本地化后形如 `1 已完成 · 2 进行中 · 1 待处理`，计数为零的段落省略），它能正确报告并行工作，且不需要任何可被截断的名字；工具行才是本分支仍需修的那一处。
+
+工具行改用 `toolviews/plan-summary.ts` 中的 `planSummary`。它给出第一个活跃条目，并计数其余活跃项，因此工具行报告的是有多少任务在跑，而不是暗示只有一个。列出全部活跃条目被否决了：工具行是单行，无上界的拼接会溢出——在列表做不到的地方，计数能够可预测地降级。该推导放在 toolviews 域内而非 `contract/`（域间共享面）：面板自行内联计算其计数，与工具行不共享任何东西，因此放进 contract 会声明一种已不存在的共享关系。
+
+`planSummary` 把任务名与计数作为两个独立字段返回，而不是一个拼好的字符串，因为工具行用 `overflow: hidden` / `text-overflow: ellipsis` 截断其摘要文本。计数接在任务名之后时位于可截断文本的末端，于是恰恰是让计数变得有意义的那些场景——窄视口、长任务名——会把它裁掉，让并行计划看起来与顺序计划无异。因此工具行把计数交给共享的 `ToolRow`，作为 `summarySuffix`——一个紧邻被省略号截断的摘要文本、且不会收缩的槽位；一个预先拼好的字符串无法表达这个切分，而把计数放到任务名之前也被否决了：读者首先要找的是任务名。
+
+`summarySuffix` 是 `ToolRow` 上的槽位，而不是 todo 工具行自有的标记：每个 toolview 都经由这个共享组件渲染，而它的 `summary` 是一个会被省略号截断的普通字符串，容不下一个必须挺过截断的片段。该后缀落在 `.summary` 规则之外，因此重复了该规则的 `font-size` 与 `line-height`——Web 外壳把正文字号留在浏览器默认值而非该行的 14px，所以未加样式的 span 会明显大于同一 24px 行内与之并列的文本。错误行会丢弃该后缀，因为它折叠态的摘要是失败行，而非任何由调用 args 推导出的内容。
+
+## 暂缓项
+
+两条 review 结论在此记录而非在本分支修复。`summarySuffix` 这个 span 没有无障碍名称，屏幕阅读器读出的数量缺少它所修饰的名词（`… 实现 fixture 样本 +1`）；为它命名会引入带自身测试契约的本地化文案，这属于对整条 `ToolRow` 摘要行做的无障碍专项，而不属于某一行。以及，当*第一个*活跃条目的 content 不可用时——缺失、类型不对、或 trim 后为空——行会连同数量一起丢掉活跃子句，于是并行计划渲染成裸计数；向后跳到第一个可用活跃条目的方案被否决，因为调用 args 是一处明确未经校验的边界，模型给出的顺序是该行唯一能遵循的顺序，而只丢掉不可用的那个子句可以保住 `done`/`total` 计数——这两个数无论如何都是可信的。
+
+## 后果
+
+现在 todo 列表可以忠实反映并行执行，并且每个展示面都能一次渲染多个活跃标记：计划横条的表头会计数活跃条目，工具行则需要上述推导。设置 `allowParallelInProgress: true` 的组合不再拒绝一种此前无效的快照形状；设置为 `false` 的组合仍保留旧的拒绝行为，而持久日志不变式两者都接受。面向模型的描述发生了变化，这重新记录了 tool-catalog 页面以及每个带有 todo schema 的快照 sidecar。此处不记录数量：该集合会随每个新落地的 pin 场景增长，而本 Note 先前记过的两次点时刻计数都在几天内失实。有效规则是：改动工具描述的分支必须刷新它分叉之后落地的那些 sidecar —— 包括固定 subagent 类工具的编号文件 `tool-schemas.<n>.expected.json`，其 schema 不被父场景覆盖 —— `pnpm run test:snapshot:refresh` 可以无 key 地对整个语料完成刷新。web fixture 的 todo 样本现在有两个条目处于 `in_progress`，因此两个由 fixture 驱动的展示面渲染的都是并行计划。`packages/client/ui-conversation/tests/todo-panel.spec.tsx` 在 src 上固定工具行摘要与计划横条，ACP `todo-write` 场景录制的是三条目、两个活跃的计划，而 `apps/web/tests/todo-row.snapshot.ts` 在组装后的应用中固定这两个面——它从构建产物 `packages/client/*/lib/client.js` 启动，因此是唯一覆盖 keyed 注册与打包接线的地方。该文件把 `summary`、`suffix` 与横条表头记录为独立字段，因此即便拼接后的文本读起来一样，把 `+N` 计数折回摘要字符串也会改变预期输出。

.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: a2c9b52507d32c2d851f811f0ecdd878a60b1e1c
-2026-07-28-tool-call-file-open-in-os.zh.md: 725db61869383711042d85cc1508b00eb1b196b6
+2026-07-28-tool-call-file-open-in-os.md: 73f5091888ab2506eab50b827e74c5120394b127
+2026-07-28-tool-call-file-open-in-os.zh.md: c1bc93c472aae8be8cee3bab6cd02f556abca494

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

@@ -12,12 +12,13 @@ Chat tool rows treated the whole summary line as a click target that opened the
 
 File-tool path summaries (`read` / `write` / `edit` args carrying `path` or `file_path`) render as hover-underline links with a pointer cursor. Clicking the path calls `host.openPath` through `WorkspacesService.openPath`, resolving relative paths against the session cwd. File-link rows disable args expand (leading icon is inert); whole-row click, row hover fill, and the click-to-open-details gesture are removed from tool rows (including bash and todo registrations). The details panel and its inject surface remain for programmatic selection; rows no longer drive them.
 
-`host.openPath` is a privileged unary RPC accepted only from loopback, same-origin browser requests (same carrier guard as `host.pickDirectory`). Platform adapters open without a shell: `open` on macOS, PowerShell `Invoke-Item` on Windows, `xdg-open` on Linux. The opener is injectable for tests. URL-only read args (`web_fetch`) are not file links.
+`host.openPath` is a privileged unary RPC accepted only from loopback, same-origin browser requests (same carrier guard as `host.pickDirectory`). Platform adapters open without a shell: `open` on macOS, PowerShell `Invoke-Item` on Windows, and `xdg-open` on desktop Linux. WSL is a separate host shape despite Node reporting `linux`: the adapter recognizes its environment or Microsoft kernel release, translates the Linux path with `wslpath -w`, and passes the resulting Windows/UNC path to the same PowerShell handoff. The opener's platform facts and command runner are injectable for tests. URL-only read args (`web_fetch`) are not file links.
 
 ## Alternatives considered
 
 - Keep row-click details and add a separate file affordance — rejected; the product ask replaces the row gesture with the file link.
 - Open files inside an in-app preview — rejected; the ask is the OS default application.
+- Treat WSL as desktop Linux — rejected; a WSL process reports `linux`, but a Linux desktop association is optional while its ordinary operator desktop and browser live on Windows.
 - Reuse `host.pickDirectory`'s timeout exemption — unnecessary; path open hand-off completes quickly under the normal unary deadline.
 
 ## Consequences
@@ -26,5 +27,5 @@ Clicking a file path in a tool row opens that path on the host. Non-file tool ro
 
 ## Risks
 
-- Linux hosts without `xdg-open` 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 row stays silent while the host returns an internal error.
 - 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

@@ -12,12 +12,13 @@ Status: implemented
 
 文件工具的路径摘要（`read`／`write`／`edit` 参数中的 `path` 或 `file_path`）渲染为悬停下划线链接并使用 pointer 光标。点击路径会经 `WorkspacesService.openPath` 调用 `host.openPath`，相对路径以会话 cwd 为基准解析。带文件链接的行关闭参数展开（左侧图标不可点）；工具行（含 bash 与 todo 注册）去掉整行点击、整行悬停底色，以及点击打开 details 的手势。details 面板及其 inject 面仍保留供程序化选择；工具行不再驱动它们。
 
-`host.openPath` 是特权一元 RPC，仅接受来自回环地址且同源的浏览器请求（与 `host.pickDirectory` 相同的载体守卫）。平台适配器不经 shell 打开：macOS 为 `open`，Windows 为 PowerShell `Invoke-Item`，Linux 为 `xdg-open`。打开器可在测试中注入。仅含 URL 的 read 参数（`web_fetch`）不是文件链接。
+`host.openPath` 是特权一元 RPC，仅接受来自回环地址且同源的浏览器请求（与 `host.pickDirectory` 相同的载体守卫）。平台适配器不经 shell 打开：macOS 为 `open`，Windows 为 PowerShell `Invoke-Item`，桌面 Linux 为 `xdg-open`。尽管 Node 将 WSL 报告为 `linux`，WSL 仍是一种独立的宿主形态：适配器根据其环境或 Microsoft 内核 release 识别它，用 `wslpath -w` 转换 Linux 路径，并将所得 Windows/UNC 路径交给同一 PowerShell 交接。打开器的平台信息和命令运行器可在测试中注入。仅含 URL 的 read 参数（`web_fetch`）不是文件链接。
 
 ## 考虑过的替代方案
 
 - 保留整行点击打开 details，另加文件入口 — 否决；产品要求用文件链接替换整行手势。
 - 在应用内预览文件 — 否决；要求是操作系统默认应用。
+- 将 WSL 当作桌面 Linux — 否决；WSL 进程报告 `linux`，但 Linux 桌面文件关联并非必有，而其常规用户桌面和浏览器位于 Windows 上。
 - 复用 `host.pickDirectory` 的超时豁免 — 不必要；打开路径的交接在常规一元截止时间内即可完成。
 
 ## 后果
@@ -26,5 +27,5 @@ Status: implemented
 
 ## 风险
 
-- 没有 `xdg-open` 的 Linux 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；聊天行保持静默，宿主返回内部错误。
 - 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/2026-07-29-persistent-bash-str-replace-editor.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/2026-07-29-persistent-bash-str-replace-editor.md
-2026-07-29-persistent-bash-str-replace-editor.md: 22851078c1cc8fa9d5716afa41c8a2e2b7e7725c
-2026-07-29-persistent-bash-str-replace-editor.zh.md: 23f80d1a5911f4f3820d526c2221d3002507d774
+2026-07-29-persistent-bash-str-replace-editor.md: c4750e30370bfd253064c39cb1adc0f5b2baa60d
+2026-07-29-persistent-bash-str-replace-editor.zh.md: 62073571ce2164d88897d4959caafd9cb1dbdd93

.agents/notes/implemented/feature/2026-07-29-persistent-bash-str-replace-editor.md

@@ -18,7 +18,7 @@ Some deployments need a one-call Bash schema whose shell state survives across m
 
 Both plugins are included in the Python runtime closure. The persistent Bash closure also includes the PTY service/local backend and the sandbox services required by that backend. Because `node-pty` executes a native `spawn-helper` on macOS, each packaged macOS runtime executable ships with a `-spawn-helper` sibling; Linux uses `forkpty` directly. A pinned `node-pty` patch checks `DSH_NODE_PTY_SPAWN_HELPER` first, so it remains a true override for a current external consumer that supplies a non-sibling helper. When the override is unset, the patch resolves the packaged executable sibling if present and otherwise preserves upstream lookup in ordinary Node runs. The macOS builders fail before publication when the helper is absent or not executable.
 
-The shipped [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) overlay composes both plugins over the ordinary Web surface, disables its other model-facing consumers, and leaves the Web host, browser, Workspace, persistence, sandbox, and permission stack in place. The local PTY backend resolves the effective session sandbox mode when it creates the shell. While that owner has an open shell or a spawn in progress, a different permission mode is rejected before its session event commits; the editor continues through the Web filesystem sandbox.
+The shipped [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) overlay composes both plugins over the ordinary Web surface for the Claude SWE-compatible RL contract. It pins native tool mode and makes the complete system prompt `DSH_SYSTEM_PROMPT` when set or `You are a helpful software engineer assistant.` otherwise, with no harness identity, source-checkout section, Web orientation, Workspace instructions, or tool-mode guidance. It disables every other model-facing consumer, so the model receives exactly the persistent `bash` and `str_replace_editor` schemas, while the Web host, browser, Workspace, persistence, sandbox, and permission stack remains in place. The local PTY backend resolves the effective session sandbox mode when it creates the shell. While that owner has an open shell or a spawn in progress, a different permission mode is rejected before its session event commits; the editor continues through the Web filesystem sandbox.
 
 ## Alternatives considered
 

.agents/notes/implemented/feature/2026-07-29-persistent-bash-str-replace-editor.zh.md

@@ -18,7 +18,7 @@ Status: implemented
 
 两个插件都进入 Python runtime 闭包。持久 Bash 的闭包还包含 PTY 服务／本地后端，以及该后端要求的沙箱服务。由于 `node-pty` 在 macOS 上会执行原生 `spawn-helper`，每个打包后的 macOS 运行时可执行文件都会携带一个 `-spawn-helper` 伴随文件；Linux 直接使用 `forkpty`。固定版本的 `node-pty` 补丁会先检查 `DSH_NODE_PTY_SPAWN_HELPER`，因此对当前提供非伴随 helper 的外部消费方而言，该变量仍是真正的覆盖项。未设置该覆盖时，补丁会在打包可执行文件的伴随文件存在时解析它，否则在普通 Node 运行中保留上游查找方式。若 helper 缺失或不可执行，macOS 构建器会在发布前失败。
 
-已交付的 [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) 覆盖层在常规 Web 界面之上组合这两个插件，禁用该界面的其他面向模型的消费方，并保留 Web 宿主、浏览器、Workspace、持久化、沙箱与权限栈。本地 PTY 后端会在创建 shell 时解析会话的有效沙箱模式。只要该所有者仍有打开的 shell 或仍在进行中的 spawn，另一种权限模式就会在对应的会话事件提交前遭到拒绝；编辑器则继续经由 Web 文件系统沙箱运行。
+已交付的 [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) overlay 会在常规 Web 界面之上组合这两个插件，以满足与 Claude SWE 兼容的 RL 契约。它固定使用原生工具模式；完整的系统提示词在设置 `DSH_SYSTEM_PROMPT` 时采用其值，否则采用 `You are a helpful software engineer assistant.`，且不包含 harness 身份、源码 checkout 提示词段、Web 界面定位、Workspace 指令或工具模式指引。它会禁用其他所有面向模型的消费方，使模型恰好只收到持久 `bash` 和 `str_replace_editor` 两个 schema，同时保留 Web 宿主、浏览器、Workspace、持久化、沙箱与权限栈。本地 PTY 后端会在创建 shell 时解析会话的有效沙箱模式。只要该所有者仍有打开的 shell 或仍在进行中的 spawn，另一种权限模式就会在对应的会话事件提交前遭到拒绝；编辑器则继续经由 Web 文件系统沙箱运行。
 
 ## 考虑过的替代方案
 

.agents/notes/implemented/feature/2026-07-29-web-message-icon-actions-and-clock.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/2026-07-29-web-message-icon-actions-and-clock.md
-2026-07-29-web-message-icon-actions-and-clock.md: f43f7f9c9687e4494993d7e225d11cf6446a9954
-2026-07-29-web-message-icon-actions-and-clock.zh.md: a6261c65c1e9d77cea2de5624b2c9fde1278c612
+2026-07-29-web-message-icon-actions-and-clock.md: feced6aeb11d176d6c774242a4d1dae14f6730f8
+2026-07-29-web-message-icon-actions-and-clock.zh.md: 5e33182421b423f45c84dbe1a979505f4c31b819

.agents/notes/implemented/feature/2026-07-29-web-message-icon-actions-and-clock.md

@@ -12,6 +12,8 @@ The web chat user bubble already had copy / branch / edit IconActions but no clo
 
 **User bubbles prepend a date-aware local clock to the existing IconActions row; the last content-text assistant of each turn appends a copy / branch / clock row with `margin-top: 16px`; both seats stay visible whenever mounted and re-format at the next local midnight.**
 
+The assistant seat is narrowed by the [completed-turn decision](../bug-fix/2026-08-05-turn-tail-actions-require-a-completed-turn.md): only a turn with a `turn/end` grants it, so a turn still producing steps hands the row to nothing. The user seat's branch control is removed outright by the [user-bubble branch removal](../simplification/2026-08-06-user-bubbles-drop-the-branch-action.md); a user row's IconActions are clock and copy.
+
 Both seats format `node.time` through `formatMessageClock`: same calendar day → `HH:mm`, earlier this year → `M月D日 HH:mm`, other years → `YYYY年M月D日 HH:mm`. `useCalendarDay` is a component-local day tick (timeout to the next local midnight) so memoized rows re-render when the calendar day changes without a new framework hook. `MessageItem` places the label before copy (figma `388:20051`). `ChatView` derives turn-tail seqs via `assistantActionsSeqs` and withholds `time` for mid-turn content; `AssistantMarkdown` places the row after branch (figma `43:32997`) only when `streaming` is false, the event time is known, and the node has non-empty text content. Think-only nodes, mid-turn narration, and the streaming tail omit the row. Copy writes joined text blocks. Both message rows pass their event's `seq` to the same fork callback; [Web session fork actions](2026-07-27-web-session-fork-actions.md) define the real mutation contract. Clipboard write and the clock helpers live in `message-chrome.ts`. The assembled surface is pinned by `apps/web/tests/message-actions.e2e.ts` (cold-seeded history + aria golden); aria normalization collapses every clock shape to `{{clock}}`.
 
 ## Alternatives considered

.agents/notes/implemented/feature/2026-07-29-web-message-icon-actions-and-clock.zh.md

@@ -12,6 +12,8 @@ Web 聊天的用户气泡已有复制、分支、编辑 IconActions，但没有
 
 **用户气泡在既有 IconActions 行的开头添加感知日期的本地时钟；每个轮次中最后一条带 text 内容的 assistant 在正文下追加带 `margin-top: 16px` 的复制、分支、时钟；两边只要挂载就保持可见，并在下一个本地午夜重新格式化。**
 
+assistant 一侧的座位由[已完成轮次决策](../bug-fix/2026-08-05-turn-tail-actions-require-a-completed-turn.md)收紧：只有存在 `turn/end` 的轮次才授予该行，仍在产出步骤的轮次不把该行交给任何节点。user 一侧的分支控件被 [user 气泡分支移除决策](../simplification/2026-08-06-user-bubbles-drop-the-branch-action.md)直接移除；user 行的 IconActions 只有时钟与复制。
+
 两边都通过 `formatMessageClock` 格式化 `node.time`：同一日历日 → `HH:mm`，同年更早 → `M月D日 HH:mm`，跨年 → `YYYY年M月D日 HH:mm`。`useCalendarDay` 是组件本地的日刻度（定时到下一个本地午夜），因此 memo 行在日历日变化时会重渲染，且不新增框架钩子。`MessageItem` 把标签放在复制之前（figma `388:20051`）。`ChatView` 通过 `assistantActionsSeqs` 推导轮次尾部的 seq，并不为轮次中间的内容传入 `time`；`AssistantMarkdown` 把该行放在分支之后（figma `43:32997`），且仅在 `streaming` 为 false、已知事件时间、且节点含非空 text 内容时渲染。纯 Think 节点、轮次中间的叙述与流式尾部省略该行。复制写入拼接后的 text 块。两种消息行都把自己的事件 `seq` 交给同一个 fork 回调；真实 mutation 契约由 [Web session fork 操作](2026-07-27-web-session-fork-actions.md)定义。剪贴板写入与时钟辅助函数放在 `message-chrome.ts`。组装后的界面由 `apps/web/tests/message-actions.e2e.ts`（冷 seed 历史 + aria golden）钉住；aria 归一化把每种时钟形态折叠为 `{{clock}}`。
 
 ## 曾考虑的方案

.agents/notes/implemented/feature/2026-07-31-code-mode-language-dispatch.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/2026-07-31-code-mode-language-dispatch.md
+2026-07-31-code-mode-language-dispatch.md: 96001252d6494d058a8df9974fb5a0d59e7d7112
+2026-07-31-code-mode-language-dispatch.zh.md: aa7eb2a6b4b9117f1d707b37afcdbe12b814bad2

.agents/notes/implemented/feature/2026-07-31-code-mode-language-dispatch.md

@@ -0,0 +1,46 @@
+# Agent Note: Code Mode language dispatch and the Python SDK renderer
+
+Status: implemented
+
+English | [中文](2026-07-31-code-mode-language-dispatch.zh.md)
+
+## Problem
+
+Code Mode generated one SDK flavor: TypeScript. `ToolRegistry` hard-coded `renderToolsSdk` for the `tools:sdk` section and `requireCodeRuntime` rejected any `ctx.codeRuntime.language !== 'typescript'`. Adding a CPython backend means a program's source language is no longer fixed: the same visible tool registry must project a Python SDK when a Python runtime is loaded, and the model-facing `run_code` schema strings ("Execute a Python program …") must match the SDK section's language so the model never sees a TypeScript instruction over a Python runtime.
+
+This is the tool-facing half of the multi-language Code Mode split; the [code-runtime seam](../../../../packages/code-runtime/code-runtime/README.md) already carries `CodeRuntime.language`. This note owns only how `dsh-tools` dispatches on that field. The backend that implements `language: 'python'` is owned by its own note, delivered separately.
+
+## Decision
+
+Language selection is a lookup on `ctx.codeRuntime.language`, resolved lazily at prompt assembly, against two parallel tables in `dsh-tools`:
+
+- `SDK_RENDERERS` (index.ts) maps a language to its `tools:sdk` renderer — `typescript → renderToolsSdk`, `python → renderToolsSdkPy`. The `tools:sdk` section reads the loaded runtime's language and picks the renderer; `requireCodeRuntime` rejects a `mode: code`/`both` runtime whose language is absent from the table, naming the known languages.
+- `RUN_CODE_FLAVORS` (code-mode.ts) maps a language to its two model-facing `run_code` strings (tool `description` and the `code` parameter description), so a language's SDK section and its transport schema always agree.
+
+Both tables are read with `Object.hasOwn` before use so a language named `toString`/`constructor` cannot resolve an inherited `Object.prototype` member as a renderer. The two guards differ in reachability: `SDK_RENDERERS`' in-callback guard is unreachable because `requireCodeRuntime` validated the same `const` table earlier in the same callback (it carries a `/* v8 ignore */`), while `RUN_CODE_FLAVORS`' guard is the primary, publicly reachable rejection — any language absent from the flavor table hits it through `run_code`'s language-aware getters, which the public `schemas()` reaches without passing `requireCodeRuntime` first; the test reads one of those getters off the definition directly, under a language absent from both tables. A language present in `SDK_RENDERERS` but not `RUN_CODE_FLAVORS` is drift the shared `CodeSdkLanguage` `satisfies` pins reject at `typecheck`, so it is not an input either guard can see; what the guards still own is a mounted runtime reporting a language absent from both tables. Schema emission reads the runtime through `peekRuntime()` rather than `requireRuntime()`: `undefined` (no runtime mounted, reached by definition readers and `schemas()`, of which the doc-catalog harvest is the only shipped one and none of which feeds a model because assembly passes `requireCodeRuntime` first) degrades to the TypeScript flavor, whereas a mounted unknown language fails loud — this is NOT the silent fallback rejected below, which concerns emitting a wrong-language SDK for a real runtime. Adding a backend language is three parallel edits — a `CodeSdkLanguage` member and the two table entries — plus its renderer and the prose that names the well-known values instead of deriving them (the seam's `dsh-code-runtime` README pair, its `CodeRuntime.language` JSDoc, and the `docs/core-data-structures/code-runtime.md` pair; this package's own README pair and its `Config.mode` JSDoc — no gate checks any of it), with no `agent-loop` or registry-structure change.
+
+`code-mode.ts` depends only on the runtime seam (`@deepseek-ai/dsh-code-runtime`), never on a concrete backend; dispatch is by `runtime.language` at run time. The tool layer therefore lands independently of the protocol and backend PRs — it needs only the seam's `language` field, which is already on master.
+
+### The Python SDK renderer
+
+`py-types.ts` renders the same unified tool-schema vocabulary `jsonSchemaToTs` covers, targeting Python: `jsonSchemaToPy` emits a type expression per JSON-schema node, and `renderToolsSdkPy` assembles named `TypedDict`s for each visible tool's arguments and canonical output plus a `tools` object with usage instructions equivalent to the TypeScript flavor. Unsupported raw constructs degrade rather than throwing during assembly, matching the TypeScript renderer's contract. The output is deterministic — lexicographic tool order, byte-identical text for an unchanged tool set — so the prompt stays prefix-cache-friendly. Lexicographic means one ordered member stream: a tool whose name is not a legal attribute is listed as a `tools[name]` comment in its sorted position rather than partitioned to the end, matching how the TypeScript flavor quotes an exotic key in place. That stream forces one thing directly — comment lines are not statements, so a tool set that emits no method at all still needs an explicit `pass`. Three further rules are Python-specific rather than consequences of the ordering. The usage contract states that the declarations are static stubs and arguments are plain `dict`/`list` values: a `TypedDict` reads as a constructible class, so a model that writes `FooArgs(field=1)` gets a `NameError` — TypeScript's `interface` is visibly a type, and the TS flavor's "runs type-stripped" clause already covers it. A description becomes the method's docstring emitted as the FIRST statement of its body: above the `async def` the first one would document the `Tools` class and the rest would be dead expressions, leaving every method undocumented. And a `list[…]` chain degrades to `Any` past `MAX_LIST_NESTING`, because CPython's tokenizer rejects a line with more than 200 open brackets and the block must stay parseable Python — the same reason `docLines` escapes quotes and backslashes. `ts-types` needs neither: TypeScript attaches a leading `/** … */` to the member that follows it and bounds nesting nowhere in its grammar.
+
+The standard that cap serves is grammatical validity, and the boundary is deliberate: a long `A | B | …` union is valid Python at any length and is left uncapped, even though CPython's `compile()` exhausts its C recursion walking the left-nested `BinOp` spine (measured on 3.9: 1,000 branches compile, 5,000 raise `RecursionError`). Nothing compiles this block — it is prompt text — so that limit costs nothing, whereas capping union length would retire the deep-chain tests that pin the walk's linear time and the class-name propagation cap. A future renderer that does need compilable output should flatten unions rather than truncate them.
+
+`renderType` validates the whole schema once (`assertSupportedJsonSchema`) and then trusts it, wrapping the walk in one `try/catch` that degrades to `Any` — the same trusted-after-validation stance the sibling `ts-types` renderer takes at this typed same-process seam ([Trust TypeScript at typed same-process seams](../../../../AGENTS.md)). It deliberately carries NO defenses against a schema whose accessors mutate between reads (post-validation cycles, TOCTOU on `const`/`enum`, self-referential functions): the input is a first-party registration (a `defineTool` literal or a raw registration) or a wire-derived plain JSON schema — the former is trusted per AGENTS.md, the latter is a `JSON.parse` product that physically cannot carry accessors, and `renderType` re-validates the whole tree on every call regardless — so such inputs are unreachable, and adding per-shape guards here would break symmetry with `ts-types` (which has none) for values the static interface forbids. `jsonSchemaToPy(schema: unknown)` accepts `unknown` and returns `Any` on a malformed schema — the Python counterpart of the TS flavor's `unknown` — but its contract is "degrade an unsupported schema", not "survive an adversarial mutating one".
+
+## Alternatives considered
+
+- **A `language` config field on `ToolRegistry`.** Deployment would then have two places to name the language (the loaded runtime and the tools config) that can disagree; the loaded runtime is the single source of truth, so the registry reads it rather than duplicating it.
+- **Importing the Python backend into `code-mode.ts` to detect it.** That would couple the tool layer to a concrete backend and force the protocol/backend PRs to land first. Runtime dispatch on `language` keeps the layer backend-agnostic and independently shippable.
+- **A default renderer for an unknown language.** A silent fallback would emit a TypeScript SDK over, e.g., a Ruby runtime — the model would see instructions in the wrong language. Failing loud at assembly is the repository's misconfiguration stance.
+
+## Consequences
+
+Adding a backend language is three parallel edits — a `CodeSdkLanguage` member, an `SDK_RENDERERS` entry, and a `RUN_CODE_FLAVORS` entry — plus the renderer function the second points at, with no change to `agent-loop` or the registry structure. The two tables (`SDK_RENDERERS`, `RUN_CODE_FLAVORS`) must stay in step, and that invariant is checked statically rather than left to review: both are `satisfies`-checked against that one union, so a language added to one and not the other fails `typecheck`. This is the mechanical form the drift risk deserves — the runtime `Object.hasOwn` guards would catch it too, but only once a backend reporting that language ships: one PR after the drift, at the consumer's integration point rather than where it was introduced, and on this base never, since no second backend exists. The tables keep their `Record<string, …>` declared type because `CodeRuntime.language` is an unconstrained `string`; the union pins what the harness ships, the guards reject what a runtime reports. What stays outside that check is the prose that names the well-known values instead of deriving them: `dsh-code-runtime`'s README pair, its `CodeRuntime.language` JSDoc, and the `docs/core-data-structures/code-runtime.md` pair at the seam, plus this package's own README pair and its `Config.mode` JSDoc. Earlier notes name the values as the state at their own PR and are not on that list. Two separate reasons keep it ungated. Prose is not type-checked at all, wherever the union lives. And no type-level pin can stand in for it here: the interface package must not import its consumer's table, and `CodeRuntime.language` stays an unconstrained `string` by design, so moving the union into the seam would not apply it either. A unit test pinning the two key sets equal was rejected in favor of this: it would buy the same check at the cost of a test-only export of two private tables, and would run later than the compiler does. Which of the two runtime failures surfaces depends on the entry point, for a language absent from both tables: assembly reports the missing renderer, because `wireSchemas` calls `requireCodeRuntime` before projecting, while the public `schemas()` reaches `run_code`'s language-aware getters first and reports the missing flavor. The tool layer stays free of any concrete backend dependency, so it lands and is testable on master ahead of the Python protocol and backend.
+
+The cost is that the Python branch of both tables is unreachable on this base: `CodeRuntime.language` is set by the loaded backend, the only published backend is `dsh-code-runtime-worker` (`'typescript'`), and the registry reads the loaded runtime rather than a config field, so no assembled application can select `renderToolsSdkPy` or `PYTHON_FLAVOR`. The model-visible surface is therefore unchanged by this note's work until a backend reporting `'python'` is published, and this PR's coverage is unit-level — the renderer output plus the dispatch and rejection paths. The keyless snapshot for the Python model interface belongs to the PR that publishes that backend, because only there does a real `cordis.yml` over published plugins produce a Python assembly; a snapshot example that mounted a fixture runtime here would assert against a test double, which [docs/testing.md](../../../../docs/testing.md) rejects as a substitute for the assembled application transcript.
+
+Two runtime contracts the Python SDK text asserts are owed by that same backend PR. First, the instructions tell the model that exactly `tools` and `ToolCallError` are bound and that the declared `TypedDict` classes are not, so the backend must inject those two names — with `ToolCallError.toolName` populated per the seam's `errorClass` contract — and must NOT bind the declared class names into the program's globals; injecting them "helpfully" would make the SDK text false. Second, the language has to be bound to the request: `requireCodeRuntime` resolves `ctx.codeRuntime` separately at assembly and at `run_code` execution, so a reload that swapped the runtime between those two points would hand a program written against one flavor to the other. The split is finer than those two points — `run_code`'s `description` and `parameters` getters each call `resolveFlavor(peekRuntime())`, and `schemaOf` destructures both, so one projection reads the runtime twice; both reads are for `run_code`'s own schema, since the getters are installed on that one definition and every other definition carries plain data properties. A reload between those two reads yields a single schema whose two halves name different languages. Neither is reachable here — one published backend means both reads return the same flavor and no program ever runs against this renderer's output — and the cross-language rejection is not testable until a second language exists.
+
+Third, that PR owns the CPython floor, and with it the renderer's Unicode-table skew. Four expressions read the running engine's tables (Node 22.23.1: Unicode 17.0) while the interpreter uses its own (CPython 3.9.6: 13.0.0): `isBareIdentifier`'s `IDENTIFIER`, and `camelCase`'s split set, head test, and `toUpperCase()`. An interpreter older than the engine is the failing direction — the engine emits a character its tokenizer refuses, taking the whole block down — and it arrives by three independent paths. Through the predicate, a bare method or field name carrying a character added between the two versions — to `XID_Start` at its head, or to `XID_Continue` in any tail position, the middle of a name included. Through `camelCase`'s XID reads, a class name, which reaches emitted text whenever any object shape in the tool's schema declares a `TypedDict`, and which the predicate's verdict on the tool name does not gate: `zz-` plus U+1E4D0 never reaches the predicate's skew, since the `-` rejects it outright, yet it still declares `class Zz𞓐xArgs`. Through the case mapping, a class name derived from a tool the predicate accepted — a different table and a wider window than XID membership: U+019B is XID_Start and NFKC-stable, so `async def ƛ` compiles on 3.9.6, but Node uppercases it to U+A7DC (unassigned there; CPython's own `.upper()` is the identity) and `class ꟜArgs` fails with `invalid non-printable character U+A7DC`. The exposure window is the characters and mappings that changed between the two versions, so the PR that names a supported CPython range must decide explicitly between accepting it and pinning all four read points to tables for that floor — pinning the predicate alone leaves both class-name paths open. Nothing here can decide it: the floor does not exist yet, and a table pinned to a guess would be a deployment-varying constant with no configurability behind it. A second axis rides along with the floor and is not one of the four: the names and syntax the block would evaluate at definition time. `TypedDict` needs 3.8, the PEP 585 builtin generics `dict[str, Any]` and `list[…]` need 3.9, an `A | B` annotation 3.10, and `NotRequired` 3.11. These are not parse failures — the block parses on any version, which is the standard the `MAX_LIST_NESTING` cap serves — but definition-time evaluation failures, and nothing in the product evaluates this text. Recording them with the read points keeps "parseable on the supported range" from being read as "executable on it".

.agents/notes/implemented/feature/2026-07-31-code-mode-language-dispatch.zh.md

@@ -0,0 +1,46 @@
+# Agent Note: Code Mode 语言分发与 Python SDK 渲染器
+
+Status: implemented
+
+[English](2026-07-31-code-mode-language-dispatch.md) | 中文
+
+## 问题
+
+Code Mode 只生成一种 SDK 形态：TypeScript。`ToolRegistry` 为 `tools:sdk` 段硬编码了 `renderToolsSdk`，且 `requireCodeRuntime` 会拒绝任何 `ctx.codeRuntime.language !== 'typescript'`。引入 CPython 后端后，程序的源语言不再固定：同一个可见工具注册表在加载 Python 运行时时必须投射出 Python SDK，而面向模型的 `run_code` schema 字符串（"Execute a Python program …"）也必须与 SDK 段的语言一致，模型才不会在 Python 运行时下看到 TypeScript 指令。
+
+这是多语言 Code Mode 拆分中面向工具的那一半；[代码运行时 seam](../../../../packages/code-runtime/code-runtime/README.md) 已经携带 `CodeRuntime.language`。本 Note 只负责 `dsh-tools` 如何在该字段上分发。实现 `language: 'python'` 的后端由它自己的 Note 负责，单独交付。
+
+## 决策
+
+语言选择就是对 `ctx.codeRuntime.language` 的查表，在 prompt 装配时惰性解析，查 `dsh-tools` 里两张平行的表：
+
+- `SDK_RENDERERS`（index.ts）把语言映射到它的 `tools:sdk` 渲染器——`typescript → renderToolsSdk`、`python → renderToolsSdkPy`。`tools:sdk` 段读取所加载运行时的语言并选出渲染器；`requireCodeRuntime` 拒绝其语言不在表中的 `mode: code`/`both` 运行时，并列出已知语言。
+- `RUN_CODE_FLAVORS`（code-mode.ts）把语言映射到它那两条面向模型的 `run_code` 字符串（工具 `description` 与 `code` 参数描述），使一种语言的 SDK 段与它的传输 schema 始终一致。
+
+两张表在使用前都以 `Object.hasOwn` 读取，这样名为 `toString`/`constructor` 的语言不会把继承自 `Object.prototype` 的成员解析成渲染器。两个守卫的可达性不同：`SDK_RENDERERS` 的段内守卫不可达，因为 `requireCodeRuntime` 已在同一回调更早处校验过同一张 `const` 表（它带 `/* v8 ignore */`）；而 `RUN_CODE_FLAVORS` 的守卫是主要的、可公开到达的拒绝路径——任何缺席 flavor 表的语言都经 `run_code` 的语言感知 getter 到达它，而公共 `schemas()` 抵达那些 getter 时并未先过 `requireCodeRuntime`；测试直读 definition 上的其中一个 getter，用的是对两张表都缺席的语言。「在 `SDK_RENDERERS` 里却不在 `RUN_CODE_FLAVORS` 里」这种漂移已由共享的 `CodeSdkLanguage` `satisfies` 在 `typecheck` 处拒绝，两个守卫都看不到这种输入；它们如今负责的是所挂载运行时报告了一门两张表都缺席的语言。schema 发射通过 `peekRuntime()` 而非 `requireRuntime()` 读取运行时：`undefined`（无运行时，由直读 definition 的读者与 `schemas()` 到达，其中 doc-catalog 采集是唯一已交付的一个，而它们都不会喂给模型，因为组装路径先过 `requireCodeRuntime`）降级到 TypeScript flavor，而挂载了未知语言则 fail loud——这不是下方被否决的静默回退，那指的是为真实运行时发出错误语言的 SDK。新增一门后端语言是三处并列编辑——一个 `CodeSdkLanguage` 成员加两条表项——再加它的渲染器，以及点名已知值而非从中派生的散文（seam 侧的 `dsh-code-runtime` README 双语对、它的 `CodeRuntime.language` JSDoc 与 `docs/core-data-structures/code-runtime.md` 双语对；本包自己的 README 双语对与它的 `Config.mode` JSDoc，无任何 gate 检查其中任何一处），不动 `agent-loop`，也不动注册表结构。
+
+`code-mode.ts` 只依赖运行时 seam（`@deepseek-ai/dsh-code-runtime`），绝不依赖具体后端；分发在运行时按 `runtime.language` 进行。因此工具层独立于协议和后端 PR 落地——它只需要 seam 的 `language` 字段，而该字段已在 master 上。
+
+### Python SDK 渲染器
+
+`py-types.ts` 渲染 `jsonSchemaToTs` 所覆盖的同一套统一工具 schema 词汇，目标为 Python：`jsonSchemaToPy` 为每个 JSON-schema 节点发出一个类型表达式，`renderToolsSdkPy` 为每个可见工具的参数与规范输出装配具名 `TypedDict`，再加一个带用法说明的 `tools` 对象，与 TypeScript 形态等价。不支持的原始构造在装配时降级而非抛错，与 TypeScript 渲染器的契约一致。输出是确定性的——工具按字典序排列，工具集不变时文本逐字节相同——因此 prompt 保持 prefix-cache 友好。字典序意味着单一有序的成员流：名字不是合法属性的工具以 `tools[name]` 注释出现在它排序后的位置上，而不是被分拣到末尾，与 TypeScript 形态就地为异常键加引号的做法一致。这个成员流直接决定了一件事：注释行不是语句，所以一个不发出任何方法的工具集仍需显式 `pass`。另有三条规则并非源自排序，而是 Python 特有。其一，用法约定声明这些声明只是静态存根、参数为普通 `dict`/`list` 值：`TypedDict` 读起来像一个可构造的类，模型若写 `FooArgs(field=1)` 会得到 `NameError`——TypeScript 的 `interface` 一眼就是类型，且 TS 形态的「runs type-stripped」一句已经覆盖了它。其二，描述会成为方法的 docstring，且必须作为方法体的**第一条语句**发出：放在 `async def` 之上，第一条会变成 `Tools` 的类文档、其余都是无效果表达式，导致每个方法都没有文档。其三，`list[…]` 链超过 `MAX_LIST_NESTING` 后降级为 `Any`，因为 CPython 的 tokenizer 拒绝一行中超过 200 个同时未闭合的括号，而这个块必须是可解析的 Python——与 `docLines` 转义引号和反斜杠是同一个理由。`ts-types` 两者都不需要：TypeScript 会把前置的 `/** … */` 附着到其后的成员上，其语法也不对嵌套设限。
+
+该上限服务的标准是**语法合法性**，这条边界是有意划定的：长的 `A | B | …` union 在任何长度下都是合法 Python，故不设上限——尽管 CPython 的 `compile()` 在沿左嵌套 `BinOp` 脊柱下降时会耗尽 C 递归（在 3.9 上实测：1,000 个分支可编译，5,000 个抛 `RecursionError`）。没有任何东西会编译这个块——它是提示词文本——所以那条限制在这里没有代价；而给 union 长度封顶会作废那几个钉住 walk 线性时间与类名传播上限的深链测试。将来若有渲染器确实需要可编译的输出，应当把 union 拍平，而不是截断。
+
+`renderType` 先用 `assertSupportedJsonSchema` 整树校验一次、随后信任它，用单个 `try/catch` 把整个遍历兜住并降级为 `Any`——与姊妹渲染器 `ts-types` 在这个 typed 同进程 seam 上采取的「校验后信任」姿态一致（[Trust TypeScript at typed same-process seams](../../../../AGENTS.md)）。它有意不设任何针对「访问器在多次读取间变值」的防御（校验后成环、`const`/`enum` 的 TOCTOU、自引用函数）：输入是第一方注册（`defineTool` 字面量或 raw 注册）或从 wire 桥接而来的纯 JSON——前者按 AGENTS.md 受信任，后者是 `JSON.parse` 产物、物理上不可能携带访问器，且每次调用 `renderType` 都会整树重新校验——这类输入不可达，而在此加逐形态守卫会为静态接口所禁止的值破坏与 `ts-types`（没有这类守卫）的对称。`jsonSchemaToPy(schema: unknown)` 接受 `unknown` 并对畸形 schema 返回 `Any`——TypeScript 形态 `unknown` 的对应物——但它的契约是「降级不支持的 schema」，而非「扛住对抗性的可变 schema」。
+
+## Alternatives considered
+
+- **在 `ToolRegistry` 上加一个 `language` 配置字段。** 那样部署方就会有两处命名语言（所加载的运行时与 tools 配置）且可能相互矛盾；所加载的运行时是唯一真相来源，故注册表读取它而不复制它。
+- **把 Python 后端 import 进 `code-mode.ts` 来检测它。** 那会把工具层耦合到具体后端，并迫使协议/后端 PR 先落地。按 `language` 运行时分发使该层保持后端无关、可独立发布。
+- **为未知语言提供默认渲染器。** 静默回退会在比如 Ruby 运行时上发出 TypeScript SDK——模型会看到错误语言的指令。在装配处 fail loud 是本仓库对错误配置的立场。
+
+## Consequences
+
+新增一门后端语言是三处并列编辑——一个 `CodeSdkLanguage` 成员、一个 `SDK_RENDERERS` 表项、一个 `RUN_CODE_FLAVORS` 表项——再加第二处所指向的渲染器函数，不动 `agent-loop`，也不动注册表结构。两张表（`SDK_RENDERERS`、`RUN_CODE_FLAVORS`）必须同步，且这条不变式由静态检查把关，而非交给 review：两张表都以 `satisfies` 对上述同一个 union 校验，因此只加其一而漏掉另一会在 `typecheck` 处失败。这正是该漂移风险应有的机械形式——运行期的 `Object.hasOwn` 守卫同样能捕获，但要等到有后端报告该语言之后：晚于漂移引入一个 PR，且触发点在消费方的集成处而非漂移引入处；在当前 base 上则永远不会触发，因为不存在第二个后端。两张表的声明类型仍是 `Record<string, …>`，因为 `CodeRuntime.language` 是不受约束的 `string`：union 钉住本仓库交付了什么，守卫拒绝运行时报告了什么。落在这条检查之外的是点名已知值而非从中派生的散文：seam 侧的 `dsh-code-runtime` README 双语对、它的 `CodeRuntime.language` JSDoc 与 `docs/core-data-structures/code-runtime.md` 双语对，再加本包自己的 README 双语对与它的 `Config.mode` JSDoc。更早的 note 点名这些值时记的是其自身 PR 当时的状态，不在此列。让它无 gate 的是两条独立理由。其一，散文根本不受类型检查，union 放在哪里都一样。其二，类型级替代在这里也不可用：接口包不得 import 其消费方的表，而 `CodeRuntime.language` 按设计保持不受约束的 `string`，即便把 union 迁进 seam 也不会作用到它。用一个断言两张表键集相等的 unit test 的方案被否决：它买到的是同一条检查，代价却是把两张私有表做测试专用导出，且运行时机晚于编译器。对两张表都缺席的语言，两种运行期失败中报出哪一条随入口而异：组装路径报缺渲染器，因为 `wireSchemas` 在投影前先调 `requireCodeRuntime`；而公共 `schemas()` 先经过 `run_code` 的语言感知 getter，报的是缺 flavor 表项。工具层不依赖任何具体后端，因此它能先于 Python 协议和后端在 master 上落地并可测。
+
+代价是两张表的 Python 分支在当前 base 上不可达：`CodeRuntime.language` 由所加载的后端设定，已发布的后端只有 `dsh-code-runtime-worker`（`'typescript'`），而注册表读取的是所加载的运行时而非某个配置字段，因此没有任何一份组装好的应用能选中 `renderToolsSdkPy` 或 `PYTHON_FLAVOR`。也就是说，在报告 `'python'` 的后端发布之前，本 note 的工作不改变模型可见表面，本 PR 的覆盖因此是 unit 级——渲染器输出加分发与拒绝路径。Python 模型界面的 keyless snapshot 归属于发布该后端的那个 PR，因为只有在那里，一份基于已发布插件的真实 `cordis.yml` 才会产出 Python 组装；在此处挂载 fixture 运行时的快照示例断言的是测试替身，而 [docs/testing.md](../../../../docs/testing.md) 明确拒绝以此替代组装好的应用 transcript。
+
+Python SDK 文本断言的两条运行时契约同样归属那个 backend PR。其一，说明文字告诉模型运行时恰好绑定 `tools` 与 `ToolCallError` 两个名字、所声明的 `TypedDict` 类不绑定，因此后端必须注入这两个名字（并按 seam 的 `errorClass` 契约填充 `ToolCallError.toolName`），且**不得**把所声明的类名绑进程序全局——「好心」注入会使这段 SDK 文本变成假话。其二，语言必须绑定到请求上：`requireCodeRuntime` 在组装时与 `run_code` 执行时分别解析 `ctx.codeRuntime`，若在这两点之间发生重载并换掉运行时，就会把针对一种形态写成的程序交给另一种形态执行。分裂比这两点更细——`run_code` 的 `description` 与 `parameters` 两个 getter 各自调用 `resolveFlavor(peekRuntime())`，而 `schemaOf` 会解构这两个字段，因此一次投影读两次运行时；两次都属于 `run_code` 自己的 schema，因为这两个 getter 只装在那一个 definition 上，其余 definition 携带的都是普通数据属性。在这两次读取之间重载会产出单个 schema 的两半分属不同语言。两者在此处都不可达——只有一个已发布后端意味着两次读取返回同一形态，且没有任何程序会针对本渲染器的输出运行——而跨语言拒绝在第二门语言存在之前也无法测试。
+
+其三，那个 PR 拥有 CPython 版本下限，连带拥有本渲染器的 Unicode 表偏斜。有四处表达式读所运行引擎的表（Node 22.23.1：Unicode 17.0），而解释器用它自己的表（CPython 3.9.6：13.0.0）：`isBareIdentifier` 的 `IDENTIFIER`，以及 `camelCase` 的切分集、头部测试与 `toUpperCase()`。解释器旧于引擎是会失败的那个方向——引擎发出的字符被其 tokenizer 拒收，整个块随之不可解析——而它经三条独立路径抵达。经判据抵达的是裸发的方法名或字段名，其中带有一个在两个版本之间新增的字符——首位加进 `XID_Start`，或尾部任意位置（含名字中部）加进 `XID_Continue`。经 `camelCase` 的 XID 读取抵达的是类名：只要工具 schema 中有任一对象形态声明 `TypedDict`，该类名就进入发出的文本，且判据对工具名的裁决并不对它设闸——工具名 `zz-` 加 U+1E4D0 因 `-` 被判据直接拒绝、从不触及那里的偏斜，却照样声明 `class Zz𞓐xArgs`。经大写映射抵达的是由判据已接受的工具派生出的类名——这是另一张表，窗口也比 XID 归属更宽：U+019B 既是 XID_Start 又 NFKC 稳定，故 `async def ƛ` 在 3.9.6 上可编译，但 Node 将其大写为 U+A7DC（在那里未分配；CPython 自己的 `.upper()` 在此是恒等），于是 `class ꟜArgs` 以 `invalid non-printable character U+A7DC` 失败。暴露窗口是两个版本之间发生变化的那些字符与映射，所以宣布支持某个 CPython 范围的那个 PR 必须在「接受该暴露」与「按该下限的表钉住全部四个读取点」之间显式作出决定——只钉判据会同时留下两条类名路径。此处无法决定：下限尚不存在，而按猜测钉死一张表会成为一个随部署而变、却没有可配置性支撑的常量。还有第二条轴随该下限一同确定，且不属于那四个读取点：本块在定义期会被求值的那些名字与语法。`TypedDict` 需要 3.8，PEP 585 的内建泛型 `dict[str, Any]` 与 `list[…]` 需要 3.9，`A | B` 形式的注解需要 3.10，`NotRequired` 需要 3.11。这些不是解析失败——本块在任何版本上都能解析，这正是 `MAX_LIST_NESTING` 上限所服务的标准——而是定义期求值失败，且产品中没有任何东西会求值这段文本。把它们与那四个读取点记在一起，可避免把「在所支持范围上可解析」读成「在其上可执行」。

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.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/2026-07-31-even-out-shipped-tool-rosters.md
-2026-07-31-even-out-shipped-tool-rosters.md: 312e61d017abad1a6ac57f2ba491a715f8fd92d0
-2026-07-31-even-out-shipped-tool-rosters.zh.md: c77370c312004052bc4f8ee9545ed287a6d92554
+2026-07-31-even-out-shipped-tool-rosters.md: d12db993654d9f2b41a16a4663dc64aefe8e3a2f
+2026-07-31-even-out-shipped-tool-rosters.zh.md: 8381a457def1a909b300de52dc011e2437c7e9f3

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.md

@@ -12,7 +12,7 @@ The result was a user-visible difference nobody had decided: the same model, ask
 
 ## Decision
 
-The rows that are not surface-specific move into [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml), and three more join them: `tool-session-query`, `tool-str-replace-editor`, and `repeat-tool-guard`. Web search moves there too; its [deployment decision](2026-07-31-web-default-search.md) owns the security boundary while the shared base owns its surface-neutral mount. Both surfaces assemble the same roster: twenty-two tools on every host — the twenty shared rows plus `glob` and `grep`, which are fixed members because `dsh-tool-fs-search` spawns the [packaged ripgrep binary](../architecture/2026-08-01-packaged-ripgrep-search.md). `tool-session-query` joined and then left again — the [session-search-not-shipped-default decision](2026-08-02-session-search-not-shipped-default.md) keeps the model-facing consumer opt-in — while the rest of this roster stands.
+The rows that are not surface-specific move into [`base.cordis.yml`](../../../../packages/bundle/base/cordis.patch.yml), and three more join them: `tool-session-query`, `tool-str-replace-editor`, and `repeat-tool-guard`. Web search moves there too; its [deployment decision](2026-07-31-web-default-search.md) owns the security boundary while the shared base owns its surface-neutral mount. Both surfaces assemble the same roster: twenty-two tools on every host — the twenty shared rows plus `glob` and `grep`, which are fixed members because `dsh-tool-fs-search` spawns the [packaged ripgrep binary](../architecture/2026-08-01-packaged-ripgrep-search.md). `tool-session-query` joined and then left again — the [session-search-not-shipped-default decision](2026-08-02-session-search-not-shipped-default.md) keeps the model-facing consumer opt-in — while the rest of this roster stands.
 
 Two rows stay surface-specific. `tmux-context` is TUI-only because a browser surface has no terminal multiplexer to describe. `session-reference` is TUI-only because it drives the shared session-query index from the launcher's process-local path, and the browser sidebar reconciles that index on its own first search.
 

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.zh.md

@@ -12,7 +12,7 @@ Status: implemented
 
 ## 决策
 
-那些并非 surface 专属的行移入 [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml),另有三行加入:`tool-session-query`、`tool-str-replace-editor` 和 `repeat-tool-guard`。Web 搜索也一并移入；其[部署决策](2026-07-31-web-default-search.md)负责安全边界，共享 base 则负责与 surface 无关的挂载。两个 surface 组装同一份清单：每台宿主上都有二十二个工具——二十个共享行加上 `glob` 和 `grep`，它们成为固定成员，因为 `dsh-tool-fs-search` 直接 spawn [打包的 ripgrep 二进制](../architecture/2026-08-01-packaged-ripgrep-search.md)。`tool-session-query` 加入后又退出了——[session-search-not-shipped-default 决策](2026-08-02-session-search-not-shipped-default.md)让面向模型的消费方保持需显式启用——而这份清单的其余部分保持不变。
+那些并非 surface 专属的行移入 [`base.cordis.yml`](../../../../packages/bundle/base/cordis.patch.yml),另有三行加入:`tool-session-query`、`tool-str-replace-editor` 和 `repeat-tool-guard`。Web 搜索也一并移入；其[部署决策](2026-07-31-web-default-search.md)负责安全边界，共享 base 则负责与 surface 无关的挂载。两个 surface 组装同一份清单：每台宿主上都有二十二个工具——二十个共享行加上 `glob` 和 `grep`，它们成为固定成员，因为 `dsh-tool-fs-search` 直接 spawn [打包的 ripgrep 二进制](../architecture/2026-08-01-packaged-ripgrep-search.md)。`tool-session-query` 加入后又退出了——[session-search-not-shipped-default 决策](2026-08-02-session-search-not-shipped-default.md)让面向模型的消费方保持需显式启用——而这份清单的其余部分保持不变。
 
 有两行仍是 surface 专属。`tmux-context` 只在 TUI,因为浏览器 surface 没有终端复用器可描述。`session-reference` 只在 TUI,因为它以 launcher 的进程本地路径驱动共享的 session-query 索引,而浏览器侧边栏会在自己的首次搜索里重建该索引。
 

.agents/notes/implemented/feature/2026-07-31-workspace-write-surface-default.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/2026-07-31-workspace-write-surface-default.md
-2026-07-31-workspace-write-surface-default.md: a0b216122e301b5332ed761155d743dc78fa3bab
-2026-07-31-workspace-write-surface-default.zh.md: 4391daa32b142ea976e3b04833163936913c17bc
+2026-07-31-workspace-write-surface-default.md: e6a36ad4ee7179cabf958114681728c3ac7340b4
+2026-07-31-workspace-write-surface-default.zh.md: 5ced928b167892d9abe1f4423da72a0243590735

.agents/notes/implemented/feature/2026-07-31-workspace-write-surface-default.md

@@ -10,7 +10,7 @@ The shipped terminal and browser surfaces exposed the same coding tools under di
 
 ## Decision
 
-[`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml) owns one sandbox and permission stack for every shipped TUI, Web, and browser-backed headless session: `dsh-sandbox-local`, `dsh-sandbox-policy`, `dsh-bash-sandbox`, `dsh-fs-sandbox`, `dsh-user-approval`, and `dsh-permission`. The composition fallback is the `workspace-write` preset, which bundles `workspace-write` file effects with the `ask` approval policy. `DSH_PERMISSION_MODE` remains an explicit process override; a stored `permission.defaultPreset` remains the user preference for later sessions and outranks the fallback through the Settings seam.
+[`base.cordis.yml`](../../../../packages/bundle/base/cordis.patch.yml) owns one sandbox and permission stack for every shipped TUI, Web, and browser-backed headless session: `dsh-sandbox-local`, `dsh-sandbox-policy`, `dsh-bash-sandbox`, `dsh-fs-sandbox`, `dsh-user-approval`, and `dsh-permission`. The composition fallback is the `workspace-write` preset, which bundles `workspace-write` file effects with the `ask` approval policy. `DSH_PERMISSION_MODE` remains an explicit process override; a stored `permission.defaultPreset` remains the user preference for later sessions and outranks the fallback through the Settings seam.
 
 A genuinely fresh session pins `permission/preset: workspace-write`, `sandbox/mode: workspace-write`, and `approval/policy: ask` before execution. Existing and resumed sessions retain their logged permission, and changing the General-settings default affects only sessions created afterward. The browser keeps its Access picker, answerable approval cards, and risk confirmation for Full access. The TUI gains the existing `/permission` command because the shared Permission service activates its command child there.
 

.agents/notes/implemented/feature/2026-07-31-workspace-write-surface-default.zh.md

@@ -10,7 +10,7 @@ Status: implemented
 
 ## 决策
 
-[`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml) 为所有已交付的 TUI、Web 以及由浏览器支撑的无头会话统一持有一套沙箱与权限栈：`dsh-sandbox-local`、`dsh-sandbox-policy`、`dsh-bash-sandbox`、`dsh-fs-sandbox`、`dsh-user-approval` 和 `dsh-permission`。组合回退值为 `workspace-write` preset，其中包含 `workspace-write` 文件效果模式与 `ask` 审批策略。`DSH_PERMISSION_MODE` 仍是显式的进程级覆盖；已存储的 `permission.defaultPreset` 仍是面向后续会话的用户偏好，并通过 Settings seam 优先于该回退值。
+[`base.cordis.yml`](../../../../packages/bundle/base/cordis.patch.yml) 为所有已交付的 TUI、Web 以及由浏览器支撑的无头会话统一持有一套沙箱与权限栈：`dsh-sandbox-local`、`dsh-sandbox-policy`、`dsh-bash-sandbox`、`dsh-fs-sandbox`、`dsh-user-approval` 和 `dsh-permission`。组合回退值为 `workspace-write` preset，其中包含 `workspace-write` 文件效果模式与 `ask` 审批策略。`DSH_PERMISSION_MODE` 仍是显式的进程级覆盖；已存储的 `permission.defaultPreset` 仍是面向后续会话的用户偏好，并通过 Settings seam 优先于该回退值。
 
 真正的新会话会在执行前固定 `permission/preset: workspace-write`、`sandbox/mode: workspace-write` 和 `approval/policy: ask`。现有会话和恢复的会话保留日志中记录的权限，更改「通用」设置中的默认值只影响之后创建的会话。浏览器保留 Access 选择器、可应答的审批卡片，以及选择 Full access 时的风险确认。共享 Permission 服务在 TUI 中激活其命令子件，因此 TUI 会获得现有的 `/permission` 命令。
 

.agents/notes/implemented/feature/2026-08-02-session-search-not-shipped-default.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/2026-08-02-session-search-not-shipped-default.md
-2026-08-02-session-search-not-shipped-default.md: ba7299712c0ba3db5e807e928f6f5d98ac917187
-2026-08-02-session-search-not-shipped-default.zh.md: 1678ebfb5514003eabe0221e460c619bab1aa444
+2026-08-02-session-search-not-shipped-default.md: 65bd72fff76210b726e7562fb8e88e5f8802434a
+2026-08-02-session-search-not-shipped-default.zh.md: 5e42a2c2323904117f9322b5c4a53c43c6ed3f2a

.agents/notes/implemented/feature/2026-08-02-session-search-not-shipped-default.md

@@ -6,11 +6,11 @@ English | [中文](2026-08-02-session-search-not-shipped-default.zh.md)
 
 ## Problem
 
-The [shipped-roster decision](2026-07-31-even-out-shipped-tool-rosters.md) made `tool-session-query` a default row of the shared [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml), so the shipped TUI and Web surfaces put the five session-search tools (`session_search`, `session_event_search`, `session_trace`, `session_event_trace`, `session_event_read`) in front of the model. That contradicted the [model-facing session-query-tools decision](2026-07-24-model-facing-session-query-tools.md), whose opt-in stance the package README recorded as "shipped host compositions do not mount it by default". The default also shipped a prompt section teaching a prior-work search workflow that no user had asked for.
+The [shipped-roster decision](2026-07-31-even-out-shipped-tool-rosters.md) made `tool-session-query` a default row of the shared [`cordis.patch.yml`](../../../../packages/bundle/base/cordis.patch.yml), so the shipped TUI and Web surfaces put the five session-search tools (`session_search`, `session_event_search`, `session_trace`, `session_event_trace`, `session_event_read`) in front of the model. That contradicted the [model-facing session-query-tools decision](2026-07-24-model-facing-session-query-tools.md), whose opt-in stance the package README recorded as "shipped host compositions do not mount it by default". The default also shipped a prompt section teaching a prior-work search workflow that no user had asked for.
 
 ## Decision
 
-The shipped TUI, Web, and headless surfaces no longer mount `@deepseek-ai/dsh-tool-session-query`: the row is removed from the shared `base.cordis.yml`, the now-dangling `disabled` patch in the opt-in [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) profile goes with it, and the workspace dependency drops from `apps/cli/package.json`. The consumer stays opt-in exactly as the model-facing-session-query-tools note describes: the ACP example's [`session-query.cordis.yml`](../../../../examples/acp-agent/session-query.cordis.yml) and its snapshot counterpart remain the mounted reference, and a custom composition can mount the package with the timeout and spill policies.
+The shipped TUI, Web, and headless surfaces no longer mount `@deepseek-ai/dsh-tool-session-query`: the row is removed from the shared `cordis.patch.yml`, the now-dangling `disabled` patch in the opt-in [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) profile goes with it, and the workspace dependency drops from `apps/cli/package.json`. The consumer stays opt-in exactly as the model-facing-session-query-tools note describes: the ACP example's [`session-query.cordis.yml`](../../../../examples/acp-agent/session-query.cordis.yml) and its snapshot counterpart remain the mounted reference, and a custom composition can mount the package with the timeout and spill policies.
 
 The `ctx.sessionQuery` service itself stays mounted. `session-query-sqlite` remains a base row — the TUI's `session-reference` consumes it for `/resume` — and the Web overlay keeps patching it to an in-memory index for the browser content search. Only the model-facing consumer is removed.
 

.agents/notes/implemented/feature/2026-08-02-session-search-not-shipped-default.zh.md

@@ -6,11 +6,11 @@ Status: implemented
 
 ## 问题
 
-[交付清单决策](2026-07-31-even-out-shipped-tool-rosters.md)把 `tool-session-query` 设为共享 [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml) 的默认行,于是交付的 TUI 与 Web surface 把这五个会话搜索工具(`session_search`、`session_event_search`、`session_trace`、`session_event_trace`、`session_event_read`)呈现给了模型。这与[面向模型的会话查询工具决策](2026-07-24-model-facing-session-query-tools.md)相抵触,该决策持需显式启用的立场,包 README 将其记录为「shipped host compositions do not mount it by default」。这份默认还交付了一个提示词段,向模型讲授一套既往工作搜索工作流,而没有任何用户要求过。
+[交付清单决策](2026-07-31-even-out-shipped-tool-rosters.md)把 `tool-session-query` 设为共享 [`cordis.patch.yml`](../../../../packages/bundle/base/cordis.patch.yml) 的默认行,于是交付的 TUI 与 Web surface 把这五个会话搜索工具(`session_search`、`session_event_search`、`session_trace`、`session_event_trace`、`session_event_read`)呈现给了模型。这与[面向模型的会话查询工具决策](2026-07-24-model-facing-session-query-tools.md)相抵触,该决策持需显式启用的立场,包 README 将其记录为「shipped host compositions do not mount it by default」。这份默认还交付了一个提示词段,向模型讲授一套既往工作搜索工作流,而没有任何用户要求过。
 
 ## 决策
 
-交付的 TUI、Web 与无头 surface 不再挂载 `@deepseek-ai/dsh-tool-session-query`:该行从共享的 `base.cordis.yml` 移除,opt-in 的 [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) profile 中那条已悬空的 `disabled` patch 也随之删除,workspace 依赖也从 `apps/cli/package.json` 中移除。该消费方仍保持 opt-in,与面向模型的会话查询工具决策所述完全一致:ACP 示例的 [`session-query.cordis.yml`](../../../../examples/acp-agent/session-query.cordis.yml) 及其快照对侧文件仍是挂载参考,自定义组合也可以连同超时与 spill 策略一起挂载该包。
+交付的 TUI、Web 与无头 surface 不再挂载 `@deepseek-ai/dsh-tool-session-query`:该行从共享的 `cordis.patch.yml` 移除,opt-in 的 [`core-web.cordis.yml`](../../../../apps/cli/config/core-web.cordis.yml) profile 中那条已悬空的 `disabled` patch 也随之删除,workspace 依赖也从 `apps/cli/package.json` 中移除。该消费方仍保持 opt-in,与面向模型的会话查询工具决策所述完全一致:ACP 示例的 [`session-query.cordis.yml`](../../../../examples/acp-agent/session-query.cordis.yml) 及其快照对侧文件仍是挂载参考,自定义组合也可以连同超时与 spill 策略一起挂载该包。
 
 `ctx.sessionQuery` 服务本身保持挂载。`session-query-sqlite` 仍是 base 的一行,TUI 的 `session-reference` 消费它来实现 `/resume`,Web overlay 也继续把它 patch 成内存索引,供浏览器内容搜索使用。被移除的只有面向模型的消费方。
 

.agents/notes/implemented/feature/2026-08-03-fs-tool-error-remedy.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/2026-08-03-fs-tool-error-remedy.md
+2026-08-03-fs-tool-error-remedy.md: f227c31365725652b130e097d70c79d3daab3684
+2026-08-03-fs-tool-error-remedy.zh.md: 11acd0cf48924833ced91591d5ea1424735969cd