Merge remote-tracking branch 'origin/master' into worktree/custom-provider-edit-parity-f8c160

2026-08-11 15:16:23 +08:00
parent 28546b67fb b6cd817aca
commit 63a294fa1b
2325 changed files with 38307 additions and 17353 deletions
@@ -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/archived/bug-fix/2026-08-10-web-favicon-dark-mode.md
+2026-08-10-web-favicon-dark-mode.md: 22e1d063a72b177e0e5c11f4bcfcbc233a81fdfc
+2026-08-10-web-favicon-dark-mode.zh.md: dff90567337c6968220091c16c6cd55cf7dad2d1
@@ -0,0 +1,26 @@
+# Agent Note: Web favicon follows the color scheme
+
+Status: implemented
+Archived: 2026-08-10
+
+English | [中文](2026-08-10-web-favicon-dark-mode.zh.md)
+
+## Problem
+
+`apps/web/public/favicon.svg` paints the DeepSeek mark solid black (`fill="#000"`), and `index.html` declares only that single SVG icon. Under an OS or browser dark color scheme the tab strip is dark too, so the black mark is effectively invisible. Safari versions before 26 do not render SVG favicons, so their users get no tab icon in any scheme.
+
+## Decision
+
+The favicon stays one file and adapts through the browser's own color-scheme signal: `favicon.svg` embeds `@media (prefers-color-scheme: dark) { path { fill: #fff } }`, switching the mark to white under a dark scheme while the light scheme keeps black. `index.html` and `manifest.webmanifest` also declare a 32×32 PNG fallback (`favicon-32x32.png`, DeepSeek brand blue `#4D6BFE`) that Safari versions before 26 render and that stays visible on both light and dark tab strips, extending the [web-install-manifest decision](../feature/2026-08-06-web-install-manifest.md).
+
+The theme signal is the OS/browser scheme, not the GUI's in-app `dsh.theme` toggle: the favicon lives in browser chrome, whose background follows the browser scheme, so `prefers-color-scheme` is the correct semantic and needs no JavaScript. Known browser quirks — Chromium may not repaint the tab icon until reload after a scheme switch, and Safari versions before 26 ignore the SVG variant — are accepted and the PNG fallback covers the older-Safari case.
+
+## Alternatives considered
+
+- **A second `<link rel="icon" media="(prefers-color-scheme: dark)">` pointing at a separate dark SVG.** Rejected: the same scheme semantics with two files to keep in sync, and no benefit over the in-file media query.
+- **A theme-presenter that swaps the icon href on `theme/change`.** Rejected: it would follow the in-app toggle rather than the browser scheme that actually colors the tab strip, and it adds client code and a presenter for a chrome asset.
+- **No PNG fallback.** Rejected: Safari versions before 26 never render SVG favicons, so the fallback is the only way those versions get a tab icon at all.
+
+## Consequences
+
+Light scheme still shows the black mark, dark scheme shows white, and Safari versions before 26 show the blue PNG in both. `apps/web/tests/pwa-manifest.e2e.ts` pins the PNG link and its order before the SVG, both manifest icons, the shipped PNG's format and dimensions, and the dark media query inside the shipped SVG. The Chromium repaint quirk remains a browser behavior the app cannot fix.
@@ -0,0 +1,26 @@
+# Agent Note: 网页图标随配色方案切换
+
+Status: implemented
+Archived: 2026-08-10
+
+[English](2026-08-10-web-favicon-dark-mode.md) | 中文
+
+## 问题
+
+`apps/web/public/favicon.svg` 把 DeepSeek 图标绘制为纯黑色（`fill="#000"`），而 `index.html` 只声明了这一个 SVG 图标。当操作系统或浏览器处于暗色配色方案时，标签栏同样是深色，黑色图标实际上不可见。Safari 26 之前的版本不渲染 SVG favicon，因此这些版本的 Safari 用户无论何种配色方案都看不到标签页图标。
+
+## 决策
+
+favicon 保持单一文件，并通过浏览器自身的配色方案信号自适应：`favicon.svg` 内嵌 `@media (prefers-color-scheme: dark) { path { fill: #fff } }`，在暗色方案下把图标切换为白色，浅色方案保持黑色。`index.html` 与 `manifest.webmanifest` 同时声明 32×32 PNG 兜底（`favicon-32x32.png`，DeepSeek 品牌蓝 `#4D6BFE`），Safari 26 之前的版本会渲染该 PNG，且它在浅色与深色标签栏上都清晰可见；这是对 [Web 安装 manifest 决策](../feature/2026-08-06-web-install-manifest.md) 的扩展。
+
+主题信号取操作系统/浏览器方案，而不是 GUI 应用内 `dsh.theme` 开关：favicon 位于浏览器 chrome 中，其背景跟随浏览器方案，因此 `prefers-color-scheme` 是正确语义，无需任何 JavaScript。已知的浏览器怪癖——Chromium 在切换方案后可能要到刷新页面才重绘标签图标，Safari 26 之前的版本忽略 SVG 变体——均被接受，旧版 Safari 场景由 PNG 兜底覆盖。
+
+## 曾考虑的替代方案
+
+- **新增指向独立暗色 SVG 的第二个 `<link rel="icon" media="(prefers-color-scheme: dark)">`。** 不予采纳：语义相同却要多维护一个文件，相比文件内媒体查询没有任何收益。
+- **由主题 presenter 在 `theme/change` 时替换图标 href。** 不予采纳：它会跟随应用内开关，而不是真正决定标签栏颜色的浏览器方案，并且为一个 chrome 资源引入客户端代码和 presenter。
+- **不提供 PNG 兜底。** 不予采纳：Safari 26 之前的版本从不渲染 SVG favicon，兜底是这些版本获得标签图标的唯一途径。
+
+## 后果
+
+浅色方案仍显示黑色图标，暗色方案显示白色，Safari 26 之前的版本两种方案都显示蓝色 PNG。`apps/web/tests/pwa-manifest.e2e.ts` 固定断言 PNG 链接及其位于 SVG 之前的顺序、manifest 中的两个图标、交付 PNG 的格式与尺寸，以及交付 SVG 内部的暗色媒体查询。Chromium 的重绘怪癖仍是浏览器行为，应用无法修复。
@@ -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/archived/feature/2026-08-08-dsh-run-headless-command.md
+2026-08-08-dsh-run-headless-command.md: ce9cff965192357022c49655983fe6ff8d554b9f
+2026-08-08-dsh-run-headless-command.zh.md: 0484c069ed6365235e4616542a4fb3d5ceb2d880
@@ -1,9 +1,12 @@
 # Agent Note: `dsh run` owns one-shot headless execution

 Status: implemented
+Archived: 2026-08-10

 English | [中文](2026-08-08-dsh-run-headless-command.zh.md)

+> **Superseded command grammar.** [Apps now own their command lines](../architecture/2026-08-06-app-owned-command-line.md): the headless startup row parses the task from `dsh --profile headless <task...>`, and the launcher no longer has a `run` invocation or patches task text into rows. This note remains the rejected launcher-owned design context; the direct execution and completion contract it selected remains current in [headless is a direct core entry point](../architecture/2026-08-09-headless-direct-core-entry-point.md).
+
 ## Problem

 Generic profile boot and one-shot task execution have different lifecycle contracts. A root grammar that accepts optional task text makes one argv shape mean either a long-lived process or a terminating task according to a plugin row discovered only after composition. It also exposes a profile implementation detail as the primary user command and gives custom profiles no explicit one-shot entry.
@@ -1,9 +1,12 @@
 # Agent Note: `dsh run` 负责一次性 headless 执行

 Status: implemented
+Archived: 2026-08-10

 [English](2026-08-08-dsh-run-headless-command.md) | 中文

+> **命令语法已被取代。** [应用现在持有自己的命令行](../architecture/2026-08-06-app-owned-command-line.md)：headless 启动行从 `dsh --profile headless <task...>` 解析任务，启动器不再包含 `run` 调用，也不再把任务文本 patch 进配置行。本笔记保留被否决的启动器持有设计背景；它选定的直接执行与完成约定仍由 [headless 是直接 core 入口](../architecture/2026-08-09-headless-direct-core-entry-point.md)持有。
+
 ## 问题

 通用 profile 启动与一次性任务执行具有不同的生命周期约定。若根语法接受可选任务文本，同一种 argv 形态会表示常驻进程或终止式任务，具体含义取决于组合完成后才发现的插件配置行。它还会把 profile 实现细节暴露成主要用户命令，并使自定义 profile 缺少明确的一次性入口。
@@ -94,6 +94,9 @@
    "bug-fix/2026-08-03-tui-long-session-render-costs.i18n.yaml": "sha256:f65f7bf8fc84c7a1f022ee393c8d969c06d9bde8bed3a0206de86fb35b246ac6",
    "bug-fix/2026-08-03-tui-long-session-render-costs.md": "sha256:6ecf2ef831f527f361ade18a882d79bc6eccf15cc676d05728e7753f41cde051",
    "bug-fix/2026-08-03-tui-long-session-render-costs.zh.md": "sha256:5f44e707b332e13fa06d625212173ea055c1c3c0aee60888435a0ff099ec6037",
+    "bug-fix/2026-08-10-web-favicon-dark-mode.i18n.yaml": "sha256:859c4399f9a017a68ba89552fdafa05e73c0599d94cee9551c84ea5b749a14f3",
+    "bug-fix/2026-08-10-web-favicon-dark-mode.md": "sha256:4d17e247abd76ae3aed5fb4e075fd66a2838292f89f7021c82a79fe37ed905e6",
+    "bug-fix/2026-08-10-web-favicon-dark-mode.zh.md": "sha256:7bbff8a3b7061c127afcc75cd2a8043b02a999b78c0180edd8f7e4807fcfe71d",
    "feature/2026-06-14-acp-agent-client-protocol.i18n.yaml": "sha256:006795baa43ae962a8d125cc0f1e9f134bc2ee9fb758b6e7669e3fa0126e1918",
    "feature/2026-06-14-acp-agent-client-protocol.md": "sha256:6828c0af74bb3fb96206ca6b21c0e56a000b50e4744aad4bc2c05092f3a5a31b",
    "feature/2026-06-14-acp-agent-client-protocol.zh.md": "sha256:ba104e841a1fb84edbd3b6c8119d50445b7785255a7a8d13bb9ac8a2cb4d2e69",
@@ -253,6 +256,9 @@
    "feature/2026-07-31-web-cards-toolrow.i18n.yaml": "sha256:f9a6ab72a77934cdcc02167c7313f08d7e9925362017b34bed7ad56c8c70fbaa",
    "feature/2026-07-31-web-cards-toolrow.md": "sha256:5058f7cec4497d1cb0a5c8e77b88fddacac6eead034f3edec88e8514919b8a3e",
    "feature/2026-07-31-web-cards-toolrow.zh.md": "sha256:ba84ef2e1be61211ab5ba6950b78ede3d3a979f252bc068d3e04e2c025f7bc03",
+    "feature/2026-08-08-dsh-run-headless-command.i18n.yaml": "sha256:1c2b4c5b61b9263b6267275d6fc69faeaad3cc887f0728a7ed4172d817af812b",
+    "feature/2026-08-08-dsh-run-headless-command.md": "sha256:7695fe7fd322377d5986f14e35f13337f4cd376405c758218a81230f6d182d1c",
+    "feature/2026-08-08-dsh-run-headless-command.zh.md": "sha256:113c14a36c64d2facc8ae46f37c7aa76359d8cacb9c18fcba26a723f15d036fb",
    "process/2026-06-11-doc-sync-enforcement.i18n.yaml": "sha256:33b6d5874427bd7a2bd82e7e2f4f482b12448b2464aef15a9c57975edb48554d",
    "process/2026-06-11-doc-sync-enforcement.md": "sha256:aa2fe83d519fc30d48dff19e596e83c8922aacc9e063e14fe2cc35b769b9100e",
    "process/2026-06-11-doc-sync-enforcement.zh.md": "sha256:698017bd35f030fdea3eac51df9e43138c48140f504739d687b7251d13fced2b",
@@ -295,12 +301,24 @@
    "process/2026-07-23-browser-demo-gif-recording.i18n.yaml": "sha256:808ccdda39e540645b440e40a2124baed737b98636265d8f6d8cf036a70f0d50",
    "process/2026-07-23-browser-demo-gif-recording.md": "sha256:4d3a3dc829c75b66f4f57a6a763b4b9562ce10efb90a19308142f598c5ea8524",
    "process/2026-07-23-browser-demo-gif-recording.zh.md": "sha256:409e5d31ea87f35c5227fcd1bd105167f580a8a24d29a223ce390b29eb31639a",
+    "process/2026-07-23-personal-staging-maintenance-skills.i18n.yaml": "sha256:cbd5c32f2997713339699f4c12876b1950895fcb21594927c7509e3cb4c5b8d5",
+    "process/2026-07-23-personal-staging-maintenance-skills.md": "sha256:a1ef960ddc47c8bc14578d8432ed0ac4272e31178937780989e4e3b3e3f4d49f",
+    "process/2026-07-23-personal-staging-maintenance-skills.zh.md": "sha256:c0f3fcf0914af88cafcc7e094a2131bd1f764003defc6eaa446b743319e73ec1",
    "process/2026-07-26-gui-pr-gif-evidence-and-assets-branch.i18n.yaml": "sha256:89b6dc255cb0dd9d97ae6f34f37d185412d4260d7bf2f87d6b25e340f01dc26e",
    "process/2026-07-26-gui-pr-gif-evidence-and-assets-branch.md": "sha256:1742e09435ade4a09349c8843eb381e870a2be2d51b74449f08260422c8096b3",
    "process/2026-07-26-gui-pr-gif-evidence-and-assets-branch.zh.md": "sha256:220bd53a88617b09ee8970627c46540fa8951724b102acfe296a7ed7d0f7b5fe",
    "process/2026-07-27-wine-windows-gates-experiment.i18n.yaml": "sha256:6f4cbc12ee9cddbb297bf7e138ccabcd204f66898a0f7411b1633f03d5a9eab5",
    "process/2026-07-27-wine-windows-gates-experiment.md": "sha256:8d37dcdab058098c7de3da1de00ce61bef92bbc8d6ee71add959474c6fb3e936",
    "process/2026-07-27-wine-windows-gates-experiment.zh.md": "sha256:77fbf04df36af09e55007a93bd6b22d08ff99869efe8de3e97dac5b4701e0a9e",
+    "process/2026-07-31-installer-adopts-existing-checkout.i18n.yaml": "sha256:e7d072856cd3df3b717d94647be4ea05db087b114df4b11d68a171677b94c970",
+    "process/2026-07-31-installer-adopts-existing-checkout.md": "sha256:d97e39d07bde30d534417a406fc6c56ee6765dbc7b671888d17ef2cac9a7c6dc",
+    "process/2026-07-31-installer-adopts-existing-checkout.zh.md": "sha256:ab445493c1aacf3f8dc091bdbd765ad56c551ae613e0d8156610f0c4614a7431",
+    "process/2026-08-04-forward-only-pr-issue-status.i18n.yaml": "sha256:af23e203a66a95674154899410e2f420d1d0685dbf856c24cfccdaa547a17925",
+    "process/2026-08-04-forward-only-pr-issue-status.md": "sha256:2d31077da47d95ab3ddf64d5efc6b1b8fb7c7709d39aca4a825ef9e9d382d501",
+    "process/2026-08-04-forward-only-pr-issue-status.zh.md": "sha256:b61f865b7a8a0ac901250a3edbb92ea73177067c4c25448c7088925c2caeccd7",
+    "process/2026-08-08-review-driven-issue-lifecycle-triggers.i18n.yaml": "sha256:4c28c59d3fc323e7cd01eff31f1fe759834719c5bede1e82b39f868970bf856d",
+    "process/2026-08-08-review-driven-issue-lifecycle-triggers.md": "sha256:1b0514de5d030170e91e12e4d6ba788a9247f840e82700faa385a1c0c76ab857",
+    "process/2026-08-08-review-driven-issue-lifecycle-triggers.zh.md": "sha256:028d78d61f603d8bac64c4cce20b393a78f8e029d3bb4976e79a47ecaefa6032",
    "simplification/2026-06-20-drop-unconsumed-llm-adapter-change-event.i18n.yaml": "sha256:ad3d1263cb0051b885173bf064de62065e2c646ccaae2d7250723da3b4eab90c",
    "simplification/2026-06-20-drop-unconsumed-llm-adapter-change-event.md": "sha256:8fb061d51c8c23b47d2367814bab3623c6d5b972f38d207a273caa9030b579bd",
    "simplification/2026-06-20-drop-unconsumed-llm-adapter-change-event.zh.md": "sha256:2ffeaca91f82844a5616d6dcce6b4af514bb8a7c46f78e47f668b204ac6edc04",
@@ -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/process/2026-07-23-personal-staging-maintenance-skills.md
-2026-07-23-personal-staging-maintenance-skills.md: a7ccc5b1e0f13e880c58a93d2e4c2cd4f06e2a93
-2026-07-23-personal-staging-maintenance-skills.zh.md: 8593291f510352429cbb29679e7d033589312e41
+2026-07-23-personal-staging-maintenance-skills.md: 86654be1afe986107e7dae4440b497a266636012
+2026-07-23-personal-staging-maintenance-skills.zh.md: 7593a2e8cafc93df02c9418c94f84bbd3f154a0c
@@ -1,6 +1,7 @@
 # Agent Note: Personal staging maintenance skills

 Status: implemented
+Archived: 2026-08-10

 English | [中文](2026-07-23-personal-staging-maintenance-skills.zh.md)

@@ -1,6 +1,7 @@
 # Agent Note: 个人集成分支维护 skill

 Status: implemented
+Archived: 2026-08-10

 [English](2026-07-23-personal-staging-maintenance-skills.md) | 中文

@@ -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/process/2026-07-31-installer-adopts-existing-checkout.md
-2026-07-31-installer-adopts-existing-checkout.md: 1d45a14273095610e3ed8047da2ce9f4ca95cbb6
-2026-07-31-installer-adopts-existing-checkout.zh.md: 971bc3b389c341b314872b8e45ab20ebd2ed5b2c
+2026-07-31-installer-adopts-existing-checkout.md: 08eec47a1cf1254409df93ab83ddf503b4d7d171
+2026-07-31-installer-adopts-existing-checkout.zh.md: dc3cc5ee8dbbe287860ad02f4bdba66afb4fbeaf
@@ -1,6 +1,7 @@
 # Agent Note: the installer adopts an existing checkout into the managed layout

 Status: implemented
+Archived: 2026-08-10

 English | [中文](2026-07-31-installer-adopts-existing-checkout.zh.md)

@@ -1,6 +1,7 @@
 # Agent Note: 安装器把已有检出接管进受管布局

 Status: implemented
+Archived: 2026-08-10

 [English](2026-07-31-installer-adopts-existing-checkout.md) | 中文

@@ -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/process/2026-08-04-forward-only-pr-issue-status.md
-2026-08-04-forward-only-pr-issue-status.md: dd567707bc7fccd0a631943ab3ffd2838a7f2f76
-2026-08-04-forward-only-pr-issue-status.zh.md: f7fee58d6afb812f97569ae4d86c3d6504f35752
+2026-08-04-forward-only-pr-issue-status.md: 56004a39ce52c77429574f481d9945cdc4936d30
+2026-08-04-forward-only-pr-issue-status.zh.md: ee85319842d3245bdfab9668de0a42ab29597fac
@@ -1,6 +1,7 @@
 # Agent Note: Forward-only PR-to-Issue status projection

 Status: implemented
+Archived: 2026-08-10

 English | [中文](2026-08-04-forward-only-pr-issue-status.zh.md)

@@ -1,6 +1,7 @@
 # Agent Note: PR 到 Issue 的状态仅向前投射

 Status: implemented
+Archived: 2026-08-10

 [English](2026-08-04-forward-only-pr-issue-status.md) | 中文

@@ -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/process/2026-08-08-review-driven-issue-lifecycle-triggers.md
-2026-08-08-review-driven-issue-lifecycle-triggers.md: 8a2d48ee23da4c20bb832ae0109e2ea9912dac83
-2026-08-08-review-driven-issue-lifecycle-triggers.zh.md: 004739ff471815b0fe12e111eba0ec7aaaef9507
+2026-08-08-review-driven-issue-lifecycle-triggers.md: 444927968912d93f473e27ae8576e8371b9c287c
+2026-08-08-review-driven-issue-lifecycle-triggers.zh.md: 6e00e2a936b6421824743e779756011fcd4a1c9e
@@ -1,6 +1,7 @@
 # Agent Note: Review-driven Issue lifecycle triggers

 Status: implemented
+Archived: 2026-08-10

 English | [中文](2026-08-08-review-driven-issue-lifecycle-triggers.zh.md)

@@ -1,6 +1,7 @@
 # Agent Note: 由评审驱动的 Issue 生命周期触发器

 Status: implemented
+Archived: 2026-08-10

 [English](2026-08-08-review-driven-issue-lifecycle-triggers.md) | 中文

@@ -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/2026-06-21-mandatory-app-attribution-headers.md
-2026-06-21-mandatory-app-attribution-headers.md: 12482ca80d19e5cd1e62b8860865db6cccea61ab
-2026-06-21-mandatory-app-attribution-headers.zh.md: 43f356887948b88d34ba41d46ed1cd2f1c89a1e2
+2026-06-21-mandatory-app-attribution-headers.md: 39050a53ec76e8c5a6cac4d8e31fa15b992c406e
+2026-06-21-mandatory-app-attribution-headers.zh.md: bfd6aa2540022f68cf9f69ccc2c12b3bc0978960
@@ -32,7 +32,7 @@ The provider-neutral identity is owned by `dsh-llm` (`packages/llm/llm/src/attri

 - product token for `User-Agent`: `deepseek-harness` (continuity with the pre-Agent Note wire value and the repo/org identity)
 - version: read from the owning package's manifest via `createRequire`, never a hand-copied constant
- app URL: `https://github.com/deepseek-ai/deepseek-harness-sdk` - the planned public home, which must exist before release
+- app URL: `https://github.com/deepseek-ai/deepseek-harness` - the repository home

 The default is mandatory and non-empty. White-label deployments pass their own `AppIdentity` to `attributionHeaders(identity)` - the override hook is the function parameter, with no deployment config plumbing until a consumer needs it - and omission falls back to the harness default rather than suppressing attribution. There is no per-request API for the model, user prompt, session id, cwd, user email, API key owner, or local machine identity to influence these fields.

@@ -77,8 +77,6 @@ The landed contract:

 **Providers see that traffic comes from the harness.** That is the point, but it means deployments that previously blended into generic SDK traffic become identifiable. Mitigation: send only static public product data and let forks/white-label deployments pass their own `AppIdentity`.

-**The app URL points at a repository that does not exist yet.** `deepseek-ai/deepseek-harness-sdk` is the planned public home; until it is created the URL is a dangling promise that blocks release.
-
 **Header support differs by client library.** The hand-rolled adapter sets headers directly; the pi-ai-backed adapter depends on pi-ai continuing to honor `StreamOptions.headers` (merged last over provider defaults). The wire-level mock-server tests are the guard: if a pi-ai upgrade stops delivering the header, the suite goes red. This is useful pressure on the abstraction: a provider adapter that cannot set mandatory headers cannot fully implement the harness LLM contract.

 **OpenRouter rankings do not benefit yet.** `User-Agent` is the correct baseline for provider-neutral HTTP identity, but it will not create OpenRouter app pages or rankings because OpenRouter requires `HTTP-Referer` for that product feature. That is deliberate: public app marketplace participation is a separate product decision, not a prerequisite for mandatory request attribution.
@@ -32,7 +32,7 @@ OpenRouter 应用归属刻意未实现。`HTTP-Referer`、`X-OpenRouter-Title`

 - `User-Agent` 的产品 token：`deepseek-harness`（与 Agent Note 之前的线路值及仓库/组织身份保持连续性）
 - 版本：通过 `createRequire` 从所属包的 manifest（元数据清单）读取，绝不手动复制常量
- 应用 URL：`https://github.com/deepseek-ai/deepseek-harness-sdk`——计划中的公开主页，且必须在发布前实际存在
+- 应用 URL：`https://github.com/deepseek-ai/deepseek-harness`——仓库主页

 默认值是强制的且非空。白标部署通过向 `attributionHeaders(identity)` 传入自己的 `AppIdentity` 来覆盖——覆盖钩子就是函数参数，在有消费方需要之前不做部署配置管道——省略时回退到 harness 默认值而非抑制归属。没有逐请求 API 允许模型、用户提示词、会话 id、cwd、用户邮箱、API key 所有者或本地机器身份影响这些字段。

@@ -77,8 +77,6 @@ OpenRouter 应用归属刻意未实现。`HTTP-Referer`、`X-OpenRouter-Title`

 **提供方看到流量来自 harness。** 这正是目的，但意味着此前混在通用 SDK 流量中的部署变得可识别。缓解措施：仅发送静态公开产品数据，并允许 fork/白标部署传入自己的 `AppIdentity`。

-**应用 URL 指向一个尚不存在的仓库。** `deepseek-ai/deepseek-harness-sdk` 是计划中的公开主页；在它创建之前，该 URL 是一个阻塞发布的悬空承诺。
-
 **不同客户端库的头部支持有差异。** 手写适配器直接设置头部；基于 pi-ai 的适配器依赖 pi-ai 继续尊重 `StreamOptions.headers`（最后合并覆盖提供方默认值）。线路级 mock 服务器测试是守卫：如果 pi-ai 升级后不再投递该头部，套件会变红。这对抽象施加了有益的压力：一个无法设置强制头部的提供方适配器不能完整实现 harness 的 LLM 约定。

 **OpenRouter 排名尚未受益。** `User-Agent` 是提供方无关的 HTTP 身份的正确基线，但它不会创建 OpenRouter 应用页面或排名，因为 OpenRouter 要求 `HTTP-Referer` 来实现该产品功能。这是有意为之：公开应用市场参与是一个独立的产品决策，不是强制请求归属的前提。
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.md
-2026-07-10-single-file-executable-sdk-runtime-distribution.md: a45678c9bb5fcae340ff7134687890879f56c630
-2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: f1fccc508471356dd6434da0e126ed38f15ed3ba
+2026-07-10-single-file-executable-sdk-runtime-distribution.md: c2b6d9ff1825915e39738bf8f782c302ecfc1d0d
+2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: d7758a77083e07b1d2cac99ae2be3f15e6edd2dc
@@ -34,27 +34,27 @@ Config discovery has two channels and fails loudly when both are missing: the `D

 ### Plugin resolution: the VFS holds a real package tree, the closure manifest IS the deploy root

-Inside the exe's VFS sits a **real package tree in build-artifact form** (each package's `lib/` plus a real `node_modules`); the Loader resolves plugin names through standard dynamic `import()`: bare specifiers resolve upward along `node_modules` from the Loader's position inside the VFS, and land inside the VFS naturally. The closed set needs no allowlist code — the set is whatever the VFS has installed, and importing a name outside the set fails.
+Inside the exe's VFS sits a **real package tree in build-artifact form** (each package's `lib/` plus a real `node_modules`). The packaged JSON-RPC entry supplies its installed harness base to app-boot's root Include: relative plugin specifiers resolve from the external configuration directory, while bare package names resolve from the VFS, so a configuration inside another Node project cannot shadow the packaged plugin set. The ordinary development bin leaves bare packages configuration-owned. Bare specifiers in the packaged entry resolve upward along `node_modules` from the entry's position inside the VFS and land inside the VFS naturally. The closed set needs no allowlist code — the set is whatever the VFS has installed, and importing a name outside the set fails.

 The deploy root is [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json) (`dsh-jsonrpc-agent-pkg`, a pnpm workspace member and a zero-code pure dependency manifest) — the unified source of truth for "which plugins the exe ships" and "what the Python runtime distributes". Adding a plugin to the exe = adding one dependency line to the manifest and repackaging. [`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) traverses every workspace package covered by that manifest and requires every non-optional workspace peer at the runtime root, reporting the complete referencing-package → missing-peer chain; `pnpm run hygiene`, CI static, and the single-exe build run it before packaging. Deploy also packs by each package's `files`, so the shared chunks tsdown splits out must be covered by `files`.

 ### Build pipeline and artifacts

-[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source, so the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory; macOS uses its target prebuild and emits the required `-spawn-helper` beside the executable. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted yields a zero-symlink file tree (most stable for the pkg VFS, physically guaranteeing a single cordis instance); disabling automatic peer installation keeps unpublished package names from triggering registry resolution; link-workspace-packages points the closure at workspace/vendor sources.
+[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → restore any direct workspace package that legacy deploy hoisted back under the source manifest's `node_modules`, omitting its package-local dependency tree and rejecting any remaining manifest gap → replace every staged dependency symlink with its target bytes, remove package-manager `.bin` links, and fail if any symlink remains → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source, so the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory; macOS uses its target prebuild and emits the required `-spawn-helper` beside the executable. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted gives pkg a stable single-instance layout that the explicit materialization pass makes symlink-free; disabling automatic peer installation prevents undeclared peers from expanding the closure; link-workspace-packages selects direct workspace dependencies. [`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) overrides the transitive `@deepseek-ai/cosmokit` and `@deepseek-ai/schemastery` semver requests to the pinned vendor sources so legacy deploy never resolves those unpublished names from a registry.

 CI: [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml), triggered explicitly only — `workflow_dispatch`, or the `build-exe` label on a pull request; native builds on the three platforms linux-x64 / linux-arm64 (`ubuntu-24.04-arm`) / macos-arm64, with `~/.pkg-cache` cached; macOS ad-hoc signing is handled by pkg. Each leg drives a mock SSE model through the SDK with the default config and a custom `cordis.yml`, drives the exe directly over NDJSON JSON-RPC, verifies the JSONL and final response, and installs release-shaped wheels into a clean venv without `runtime_bin`; Linux additionally inspects GLIBC requirements and runs in a manylinux 2.28 container. A full three-target run retains four artifacts, each containing one release file: the platform-independent SDK wheel and three native runtime wheels; a subset dispatch retains the SDK wheel and selected runtime wheels. Bare executables and source bundles remain intermediate test inputs. [`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) accepts only `python-vX.Y.Z` tag pipelines whose version matches the root `package.json`, builds one SDK wheel and three native runtime wheels, then a single serialized job checks and publishes all four to the project PyPI registry. Windows is a non-goal.

 ### Python SDK distribution: two carriers, exe for production, node for development

-The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe and optional helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.
+The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe and optional helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.

-[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) reads the authoritative stable `X.Y.Z` from the repository root `package.json` and stages both packages at that version, with the SDK depending exactly on `deepseek-harness-runtime-bin==X.Y.Z`. An optional `python-vX.Y.Z` release tag is a consistency assertion and is rejected when it differs from the repository version; the source `pyproject.toml` development sentinel never determines a release version. The SDK is a `py3-none-any` wheel; each wheel-only runtime package contains one exe, and the macOS wheel also contains its architecture-matched helper. Runtime wheels use one of `py3-none-manylinux_2_28_x86_64`, `py3-none-manylinux_2_28_aarch64`, or `py3-none-macosx_11_0_arm64`; the Hatch hook rejects sdists, universal tags, mixed-platform payloads, missing or extra helpers, and unsupported platforms.
+[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) reads the authoritative stable `X.Y.Z` from the repository root `package.json` and stages both packages at that version, with `deepseek-harness-sdk` depending exactly on `deepseek-harness-runtime-bin==X.Y.Z`. An optional `python-vX.Y.Z` release tag is a consistency assertion and is rejected when it differs from the repository version; the source `pyproject.toml` development sentinel never determines a release version. The SDK is a `py3-none-any` wheel; each wheel-only runtime package contains one exe, and the macOS wheel also contains its architecture-matched helper. Runtime wheels use one of `py3-none-manylinux_2_28_x86_64`, `py3-none-manylinux_2_28_aarch64`, or `py3-none-macosx_11_0_arm64`; the Hatch hook rejects sdists, universal tags, mixed-platform payloads, missing or extra helpers, and unsupported platforms.

 The exe's "must be explicitly configured" hard semantic is unchanged; the zero-config experience is restored by the wrapper: when the caller gave no `cordis`, named no explicit runtime, and the environment has no `DSH_CORDIS_CONFIG`, the client explicitly injects the checked-in default `cordis.yml` (agent-core + preloaded llm-deepseek + JSONL persistence + bash-local + the `dsh-jsonrpc` serving entry, with `!!js` environment-variable fallbacks) via `DSH_CORDIS_CONFIG`.

 ### Naming lineage

-`@deepseek-ai/dsh-jsonrpc-demo` (the package) → `dsh-jsonrpc-agent` (the bin) → `dsh-jsonrpc-agent-pkg` (the closure manifest; no scope prefix, deliberately sidestepping the constraints' package-shape rules for `@deepseek-ai/dsh-*`) → `dsh-jsonrpc-agent-pkg-<platform>-<arch>` (the exe artifacts). The wire `serverInfo.name` stays `deepseek-harness-sdk-runtime` (a protocol-stable value); the Python dist names are `deepseek-harness` / `deepseek-harness-runtime-bin`.
+`@deepseek-ai/dsh-jsonrpc-demo` (the package) → `dsh-jsonrpc-agent` (the bin) → `dsh-jsonrpc-agent-pkg` (the closure manifest; no scope prefix, deliberately sidestepping the constraints' package-shape rules for `@deepseek-ai/dsh-*`) → `dsh-jsonrpc-agent-pkg-<platform>-<arch>` (the exe artifacts). The wire `serverInfo.name` stays `deepseek-harness-sdk-runtime` (a protocol-stable value); the Python distribution names are `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`, while the import modules remain `deepseek_harness` / `deepseek_harness_runtime`.

 ## Disposition of worker-style plugins

@@ -62,7 +62,7 @@ The exe's "must be explicitly configured" hard semantic is unchanged; the zero-c

 ## Testing

-The verification surface has three tiers. Mechanism tier: the measured conclusions for the `--sea` chain are embedded in the Decision sections (ESM dynamic import inside the VFS, single cordis instance, fail-loud config chain, `node:sqlite`, macOS ad-hoc signing runs). SDK tier: the complete keyless pytest suite covers the client protocol against a fake runtime peer, subprocess cleanup, absolute cwd propagation, dual-carrier launch, and carrier resolution; root CI runs it on Python 3.10. End-to-end tier: every platform build completes a turn against a mock endpoint through the default SDK path, a custom config, and the direct binary protocol, with final text and JSONL checked. The custom config additionally drives `run_code` and a zero-agent `workflow` through their real worker files inside the packaged VFS. The same build leg runs a committed executable-specific snapshot through the Python SDK: a keyless scripted model mounts a Cordis plugin that registers a tool, invokes that tool from `run_code`, runs a direct spawn subagent and a workflow that starts a second spawn child, then unmounts the plugin. The fixture explicitly disables its unused bundled Bash and local skill discovery so its tool set does not depend on repository-external state, and the comparison normalizes opaque message IDs in the SDK result and notification stream plus the parent and two child JSONL logs. This harness stays separate from ACP's `pnpm run test:snapshot` because the protocols and build artifacts differ. The platform wheel is then installed in a clean venv and run without `runtime_bin`.
+The verification surface has three tiers. Mechanism tier: the measured conclusions for the `--sea` chain are embedded in the Decision sections (ESM dynamic import inside the VFS, single cordis instance, fail-loud config chain, `node:sqlite`, macOS ad-hoc signing runs). SDK tier: the complete keyless pytest suite covers the client protocol against a fake runtime peer, subprocess cleanup, absolute cwd propagation, dual-carrier launch, and carrier resolution; root CI runs it on Python 3.10. End-to-end tier: every platform build completes a turn against a mock endpoint through the default SDK path, a custom config, the checked-in standalone minimal composition, and the direct binary protocol, with final text and JSONL checked. The minimal run asserts its exact system prompt and two-tool catalog, retains Bash state across calls, and invokes the editor. The custom config additionally drives `run_code` and a zero-agent `workflow` through their real worker files inside the packaged VFS. The same build leg runs a committed executable-specific snapshot through the Python SDK: a keyless scripted model mounts a Cordis plugin that registers a tool, invokes that tool from `run_code`, runs a direct spawn subagent and a workflow that starts a second spawn child, then unmounts the plugin. The fixture explicitly disables its unused bundled Bash and local skill discovery so its tool set does not depend on repository-external state, and the comparison normalizes opaque message IDs in the SDK result and notification stream plus the parent and two child JSONL logs. This harness stays separate from ACP's `pnpm run test:snapshot` because the protocols and build artifacts differ. The platform wheel is then installed in a clean venv and run without `runtime_bin`.

 Manual-driving caveat: the bin treats stdin EOF as "the client is gone" and disposes immediately, so a short-lived pipe aborts an in-flight turn — pipe-driven runs must keep stdin open until the turn ends.

@@ -34,27 +34,27 @@ exe 使用 [@yao-pkg/pkg](https://github.com/yao-pkg/pkg)（vercel/pkg 归档后

 ### 插件解析：VFS 装载真实包树，闭包 manifest（元数据清单）就是部署根目录

-exe 的 VFS 内是**构建产物形态的真实包树**（各包的 `lib/` + 真实 `node_modules`）。loader 通过标准动态 `import()` 解析插件名：裸包名从 VFS 内 loader 所在位置沿 `node_modules` 向上解析，自然落在 VFS 内。封闭集不需要白名单代码——VFS 中安装了什么，集合中就有什么；`import()` 集合外的名称会失败。
+exe 的 VFS 内是**构建产物形态的真实包树**（各包的 `lib/` + 真实 `node_modules`）。打包专用 JSON-RPC 入口会向 app-boot 的根 Include 提供自身已安装 harness 的基准位置：相对插件说明符从外部配置目录解析，裸包名则从 VFS 解析，因此位于另一个 Node 项目内的配置无法遮蔽已打包的插件集合。普通开发 bin 仍由配置项目提供裸包。打包入口中的裸包名从该入口在 VFS 内的位置沿 `node_modules` 向上解析，自然落在 VFS 内。封闭集不需要白名单代码——VFS 中安装了什么，集合中就有什么；`import()` 集合外的名称会失败。

 部署根目录是 [`python/sdk-runtime/package.json`](../../../../python/sdk-runtime/package.json)（`dsh-jsonrpc-agent-pkg`，pnpm 工作区成员、零代码纯依赖 manifest），也是「exe 安装哪些插件」与「Python 运行时分发什么」的统一真源。向 exe 添加插件，就是在 manifest 中增加一行依赖后重新打包。[`scripts/verify-runtime-closure.ts`](../../../../scripts/verify-runtime-closure.ts) 遍历该 manifest 覆盖的全部工作区包，要求每个非可选的工作区对等依赖（peer dependency）都显式列在运行时根目录，并报告“引用包 → 缺失对等依赖”的完整链路；`pnpm run hygiene`、CI 静态检查与 single-exe 构建都会在打包前运行该门禁。部署还会依据各包的 `files` 字段打包，因此 tsdown 拆出的共享分片必须被 `files` 覆盖。

 ### 构建管线与产物

-[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`，而 `--legacy` 部署会省略该副作用目录，因此构建器会把它从根安装目录复制到暂存闭包；macOS 使用对应目标的预构建产物，并在可执行文件旁生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 产出无符号链接的文件树（对 pkg VFS 最稳定，并从物理上保证只有一个 Cordis 实例）；关闭对等依赖自动安装可避免未发布包名触发注册表解析；`link-workspace-packages` 让闭包指向工作区/vendor 源码。
+[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 恢复被 legacy deploy 提升回源 manifest 的 `node_modules` 下的任何直接工作区包，同时省略其包内依赖树，并拒绝剩余的 manifest 缺口 → 将暂存依赖中的每个符号链接替换为目标文件内容，删除包管理器的 `.bin` 链接，并在仍有任何符号链接时失败 → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`，而 `--legacy` 部署会省略该副作用目录，因此构建器会把它从根安装目录复制到暂存闭包；macOS 使用对应目标的预构建产物，并在可执行文件旁生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 为 pkg 提供稳定的单实例布局，再由显式物化步骤消除符号链接；关闭对等依赖自动安装可防止未声明的对等依赖扩大闭包；`link-workspace-packages` 选择直接工作区依赖。[`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) 将传递的 `@deepseek-ai/cosmokit` 与 `@deepseek-ai/schemastery` semver 请求覆盖到固定的 vendor 源码，使 legacy deploy 不会从注册表解析这些未发布名称。

 CI 使用 [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml)，且只允许显式触发：手动派发 `workflow_dispatch`，或给 PR（Pull Request）添加 `build-exe` 标签。linux-x64、linux-arm64（`ubuntu-24.04-arm`）和 macos-arm64 三个平台分别进行原生构建，并缓存 `~/.pkg-cache`；macOS 的 ad-hoc 签名由 pkg 处理。每个平台都使用 mock SSE（Server-Sent Events）模型，分别通过默认配置和自定义 `cordis.yml` 驱动 SDK，再通过 NDJSON JSON-RPC 直接驱动 exe，校验 JSONL 与最终响应；最后把发布形态的 wheel 包安装到干净的 venv 中，并在不传 `runtime_bin` 的情况下运行。Linux 还会检查 GLIBC 依赖，并在 manylinux 2.28 容器中运行。完整构建三个目标时保留 4 个产物，每个产物只含一个发布文件：平台无关的 SDK wheel 包与 3 个原生运行时 wheel 包；手动选择部分目标时保留 SDK wheel 与所选运行时 wheel。裸 exe 与源码包只作为测试中间输入。[`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) 只接受版本与根目录 `package.json` 匹配的 `python-vX.Y.Z` 标签流水线，构建一个 SDK wheel 包和 3 个原生运行时 wheel 包，再由单个串行任务校验并将这 4 个文件发布到项目的 PyPI 注册表。Windows 不在目标范围内。

 ### Python SDK 分发：双载体，exe 用于生产，`node` 用于开发

-Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 与可选 helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。
+Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 与可选 helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。

-[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) 从仓库根目录的 `package.json` 读取权威的稳定版本 `X.Y.Z`，以该版本暂存两个包，并让 SDK 精确依赖 `deepseek-harness-runtime-bin==X.Y.Z`。可选的 `python-vX.Y.Z` 发布标签只是一项一致性断言，与仓库版本不同时会被拒绝；源码 `pyproject.toml` 中的开发占位版本从不决定发布版本。SDK 是 `py3-none-any` wheel 包；每个只提供 wheel 包的运行时包都包含一个 exe，macOS wheel 包还包含与其架构匹配的 helper。运行时 wheel 包使用 `py3-none-manylinux_2_28_x86_64`、`py3-none-manylinux_2_28_aarch64` 或 `py3-none-macosx_11_0_arm64` 三种标签之一；Hatch 钩子拒绝 sdist、通用标签、混合平台载荷、helper 缺失或多余，以及不支持的平台。
+[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) 从仓库根目录的 `package.json` 读取权威的稳定版本 `X.Y.Z`，以该版本暂存两个包，并让 `deepseek-harness-sdk` 精确依赖 `deepseek-harness-runtime-bin==X.Y.Z`。可选的 `python-vX.Y.Z` 发布标签只是一项一致性断言，与仓库版本不同时会被拒绝；源码 `pyproject.toml` 中的开发占位版本从不决定发布版本。SDK 是 `py3-none-any` wheel 包；每个只提供 wheel 包的运行时包都包含一个 exe，macOS wheel 包还包含与其架构匹配的 helper。运行时 wheel 包使用 `py3-none-manylinux_2_28_x86_64`、`py3-none-manylinux_2_28_aarch64` 或 `py3-none-macosx_11_0_arm64` 三种标签之一；Hatch 钩子拒绝 sdist、通用标签、混合平台载荷、helper 缺失或多余，以及不支持的平台。

 exe「必须显式配置」的硬语义不变；零配置体验由包装层恢复：调用方没有提供 `cordis`、没有显式指定运行时，且环境中没有 `DSH_CORDIS_CONFIG` 时，客户端将检入的默认 `cordis.yml`（`agent-core` + 预载的 `llm-deepseek` + JSONL 持久化 + `bash-local` + `dsh-jsonrpc` 对外服务条目，并通过 `!!js` 使用环境变量兜底）显式注入 `DSH_CORDIS_CONFIG`。

 ### 命名血统

-`@deepseek-ai/dsh-jsonrpc-demo`（包）→ `dsh-jsonrpc-agent`（`bin`）→ `dsh-jsonrpc-agent-pkg`（闭包 manifest；没有作用域前缀，刻意避开 `constraints` 对 `@deepseek-ai/dsh-*` 的包形状规则）→ `dsh-jsonrpc-agent-pkg-<platform>-<arch>`（exe 产物）。协议字段 `serverInfo.name` 保持为 `deepseek-harness-sdk-runtime`（协议稳定值）；Python 分发名为 `deepseek-harness` / `deepseek-harness-runtime-bin`。
+`@deepseek-ai/dsh-jsonrpc-demo`（包）→ `dsh-jsonrpc-agent`（`bin`）→ `dsh-jsonrpc-agent-pkg`（闭包 manifest；没有作用域前缀，刻意避开 `constraints` 对 `@deepseek-ai/dsh-*` 的包形状规则）→ `dsh-jsonrpc-agent-pkg-<platform>-<arch>`（exe 产物）。协议字段 `serverInfo.name` 保持为 `deepseek-harness-sdk-runtime`（协议稳定值）；Python 分发包名为 `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`，导入模块名仍为 `deepseek_harness` / `deepseek_harness_runtime`。

 ## 工作线程插件

@@ -62,7 +62,7 @@ exe 内支持 `dsh-workflow-workerthread` 与 `dsh-code-runtime-worker`。两个

 ## 测试

-验证面分三层。机制层：`--sea` 链路的实测结论内嵌在「决策」各节（VFS 内 ESM 动态 `import()`、单一 Cordis 实例、明确报错的配置链路、`node:sqlite`、macOS ad-hoc 签名可运行）。SDK 层：完整的无密钥 pytest 套件以 mock 运行时对端覆盖客户端协议、子进程清理、绝对 `cwd` 传递、双载体启动与载体解析；根 CI 在 Python 3.10 上运行全部用例。端到端层：每个平台构建都通过默认 SDK 路径、自定义配置和直接二进制协议，对 mock 端点完成一个轮次，并校验最终文本与 JSONL。自定义配置还会通过打包进 VFS 的真实工作线程文件执行 `run_code` 和不启动 agent 的 `workflow`。同一构建任务还会经 Python SDK 运行一组检入的 exe 专用快照：无密钥脚本化模型挂载一个会注册工具的 Cordis 插件，从 `run_code` 调用该工具，运行一个直接 spawn 的 subagent 和一个会通过 spawn 启动第二个 subagent 的工作流，随后卸载该插件。该 fixture（测试前置数据）会显式禁用组合包中未使用的 Bash 和本地 skill（技能）发现，使其工具集不依赖仓库外部状态；比较时会规范化以下各处的不透明消息 ID：SDK 结果与通知流，以及父会话和两个子会话的 JSONL 日志。该 harness 与 ACP 的 `pnpm run test:snapshot` 保持独立，因为二者的协议和构建产物不同。随后把平台 wheel 包安装进干净的 venv，并在不传 `runtime_bin` 的情况下运行。
+验证面分三层。机制层：`--sea` 链路的实测结论内嵌在「决策」各节（VFS 内 ESM 动态 `import()`、单一 Cordis 实例、明确报错的配置链路、`node:sqlite`、macOS ad-hoc 签名可运行）。SDK 层：完整的无密钥 pytest 套件以 mock 运行时对端覆盖客户端协议、子进程清理、绝对 `cwd` 传递、双载体启动与载体解析；根 CI 在 Python 3.10 上运行全部用例。端到端层：每个平台构建都通过默认 SDK 路径、自定义配置、仓库内置的独立 minimal 组合和直接二进制协议，对 mock 端点完成一个轮次，并校验最终文本与 JSONL。minimal 运行会断言其精确系统提示词与双工具目录，跨调用保留 Bash 状态，并调用编辑器。自定义配置还会通过打包进 VFS 的真实工作线程文件执行 `run_code` 和不启动 agent 的 `workflow`。同一构建任务还会经 Python SDK 运行一组检入的 exe 专用快照：无密钥脚本化模型挂载一个会注册工具的 Cordis 插件，从 `run_code` 调用该工具，运行一个直接 spawn 的 subagent 和一个会通过 spawn 启动第二个 subagent 的工作流，随后卸载该插件。该 fixture（测试前置数据）会显式禁用组合包中未使用的 Bash 和本地 skill（技能）发现，使其工具集不依赖仓库外部状态；比较时会规范化以下各处的不透明消息 ID：SDK 结果与通知流，以及父会话和两个子会话的 JSONL 日志。该 harness 与 ACP 的 `pnpm run test:snapshot` 保持独立，因为二者的协议和构建产物不同。随后把平台 wheel 包安装进干净的 venv，并在不传 `runtime_bin` 的情况下运行。

 手工驱动注意：`bin` 将 stdin EOF 视为「客户端已离开」并立即 dispose，短命管道会中止进行中的轮次——管道驱动必须保持 stdin 打开，直到轮次结束。

@@ -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-gui-layering-and-rpc-protocol.md
-2026-07-19-gui-layering-and-rpc-protocol.md: 514deb890d4e08d465db869669078473d32fb215
-2026-07-19-gui-layering-and-rpc-protocol.zh.md: f6fa71e3dac25f48b2ad4744a0cc695417528b34
+2026-07-19-gui-layering-and-rpc-protocol.md: da96ae97f2a2d64aeef7794bd82ccbd86602b1ad
+2026-07-19-gui-layering-and-rpc-protocol.zh.md: 36dc7391bc3f9bb0d5105fea14a2763d0b7159a1
@@ -10,7 +10,7 @@ English | [中文](2026-07-19-gui-layering-and-rpc-protocol.zh.md)

 We need a UI integration layer. Beyond the existing ACP/stdio baseline, more product clients are coming — Web (server), Electron, and others. We call them Clients and want the following capabilities:

- One `dsh` process supporting both `dsh web` (serve) and `dsh run` (headless) — one process, two modes (a design reservation)
+- One `dsh` process supporting both `dsh web` (serve) and `dsh --profile headless` (headless) — one process, two modes (a design reservation)
 - Launching inside Electron with the same Web technologies as `dsh web`

 That demands a stable layered responsibility model in the engineering codebase, so future clients plug in cleanly.
@@ -31,7 +31,7 @@ Directories layer as follows:
    - **Fetch-arrival plugin packages** (`ui-layout`, `ui-sidebar`, `ui-conversation`, `ui-trajectory`): dual-entry — the root index is the node half (an empty `apply`, existing so the host Loader governs lifecycle and the web plugin registry discovers the package.json `dsh.client` declaration); the implementation lives under `src/client/`, shipped as the `./client` subpath (a tsdown closure-factory bundle). Cross-plugin consumption of `/client` is type-only; value cooperation goes through cordis services.
 - `apps/` holds the externally exported applications, assembled from Client / Host mixtures.
    - `apps/web` (`dsh-frontend`) is the vite application: a thin `main.ts` over the shell surface exported by `dsh-client-web`.
-    - `apps/cli` (`@deepseek-ai/dsh`) dispatches commands: `dsh web` = Host + webserver + the built `dsh-frontend` dist; `dsh run` = [a direct core Agent/Session entry point](2026-08-09-headless-direct-core-entry-point.md), with zero Host, HTTP, or browser layer.
+    - `apps/cli` (`@deepseek-ai/dsh`) dispatches commands: `dsh web` = Host + webserver + the built `dsh-frontend` dist; `dsh --profile headless` = [a direct core Agent/Session entry point](2026-08-09-headless-direct-core-entry-point.md), with zero Host, HTTP, or browser layer.
    - A future Electron application reuses the same web client packages over an IPC fetch carrier.

 ```
@@ -79,7 +79,7 @@ Packages under `packages/host/*` and `packages/client/*` **must carry the direct
 2. **Write an assembly module under `apps/`**: `startHost()` + a client subclass + the application's private signal/print/exit semantics; a mixture never becomes a package — assembly is written in the app.
 3. **Import `dsh-host-webserver` only if you need HTTP carriage**, otherwise zero ports.

-The two existing applications preserve the division: the Web application mounts Host, carrier, and browser composition, while `dsh run` mounts a direct core runner with zero Host, HTTP, or ports. ACP-class protocol bridges do not follow the client-carrier checklist: they expose core to the external ecosystem and mount directly via `ctx.plugin(entry-point plugin)` without fetch.
+The two existing applications preserve the division: the Web application mounts Host, carrier, and browser composition, while `dsh --profile headless` mounts a direct core runner with zero Host, HTTP, or ports. ACP-class protocol bridges do not follow the client-carrier checklist: they expose core to the external ecosystem and mount directly via `ctx.plugin(entry-point plugin)` without fetch.

 ## Message protocol

@@ -215,7 +215,7 @@ All four quadrant full forms pass through `onEnvelope`; the base implementation

 | Subclass | Package | doFetch | Purpose |
 |---|---|---|---|
-| `InProcessApiClient` | apiproxy itself | the injected `{ fetch }` handler | **The isomorphic point**: `new InProcessApiClient(toFetchHandler(api))` never touches the network yet runs the real wire serialization/zod/SSE framing; carrier tests and callers can exercise the protocol without opening a port, while product `dsh run` drives core directly |
+| `InProcessApiClient` | apiproxy itself | the injected `{ fetch }` handler | **The isomorphic point**: `new InProcessApiClient(toFetchHandler(api))` never touches the network yet runs the real wire serialization/zod/SSE framing; carrier tests and callers can exercise the protocol without opening a port, while product `dsh --profile headless` drives core directly |
 | `WebApiClient` | dsh-client-connection | `globalThis.fetch` uplink + one same-origin WebSocket downlink per logical stream | the browser client; physical boundary in the [WebSocket downlink carrier](2026-08-04-websocket-downlink-carrier.md) |
 | `FixtureApiClient` | dsh-client-connection | unused (protocol-layer override) | serverless UI development (`?fixture`): overrides the `callUnary`/`openMux`/`openHost`/`respond` virtuals and is itself the fake server (frame rpcIds minted by it, semantics self-consistent) |
 | IPC bridge subclass (hypothetical example — no such shell exists) | an Electron shell | IPC serialization round trip | would swap only doFetch; contract and base class unchanged |
@@ -9,7 +9,7 @@ Status: implemented
 ## Problem

 需要提供 UI 对接层，除已有 ACP（Agent Client Protocol）/stdio 基线外，还需要 Web（server）、Electron 等其他产品客户端。我们把它们统一称为 Client。希望具备以下能力：
- 一个 `dsh` 进程同时支持 `dsh web`（启动）和 `dsh run`（headless），一个进程两种模式（设计预留）
+- 一个 `dsh` 进程同时支持 `dsh web`（启动）和 `dsh --profile headless`（headless），一个进程两种模式（设计预留）
 - 在 Electron 中使用与 `dsh web` 相同的 Web 技术启动

 那么当前的工程代码需要稳定的分层职责模型，便于以后接入各类 client。
@@ -29,7 +29,7 @@ Status: implemented
    - **fetch 到达插件包**（`ui-layout`、`ui-sidebar`、`ui-conversation`、`ui-trajectory`）：双入口——根入口是 node 半边（空 `apply`，其存在是为了让 host Loader 管辖生命周期、让 web 插件注册表发现 package.json 的 `dsh.client` 声明）；实现住在 `src/client/` 下，经 `./client` 子路径发布（tsdown 闭包工厂 bundle）。跨插件消费 `/client` 只限类型；值层面的协作走 cordis 服务。
 - `apps/` 作为对外导出的应用入口，可以由 Client / Host 混合组装。
    - `apps/web`（`dsh-frontend`）是 vite 应用：`dsh-client-web` 导出的壳表面之上的一层薄 `main.ts`。
-    - `apps/cli`（`@deepseek-ai/dsh`）分发命令：`dsh web` = Host + webserver + 构建出的 `dsh-frontend` dist；`dsh run` = [直接使用核心 Agent／Session 的入口](2026-08-09-headless-direct-core-entry-point.md)，不含 Host、HTTP 或浏览器层。
+    - `apps/cli`（`@deepseek-ai/dsh`）分发命令：`dsh web` = Host + webserver + 构建出的 `dsh-frontend` dist；`dsh --profile headless` = [直接使用核心 Agent／Session 的入口](2026-08-09-headless-direct-core-entry-point.md)，不含 Host、HTTP 或浏览器层。
    - 将来的 Electron 应用经由 IPC fetch 载体复用同一套 web client 包。

 ```
@@ -77,7 +77,7 @@ TypeScript 以 solution 根引用的**两个聚合 program** 检查（`tsconfig.
 2. **在 `apps/` 下写拼装模块**：`startHost()` + 客户端子类 + 该应用私有的信号/打印/退出语义；混合体不建包，拼装写在 app 里。
 3. **需要 HTTP 承载才 import `dsh-host-webserver`**，否则零端口。

-现有两个应用保持这一区分：Web 应用挂载 Host、载体与浏览器组合，而 `dsh run` 挂载直接使用核心服务的 runner，不包含 Host、HTTP 或端口。ACP 类协议桥不遵循 client 载体清单：它把 core 暴露给外部生态，直接通过 `ctx.plugin(入口插件)` 挂载，不使用 fetch。
+现有两个应用保持这一区分：Web 应用挂载 Host、载体与浏览器组合，而 `dsh --profile headless` 挂载直接使用核心服务的 runner，不包含 Host、HTTP 或端口。ACP 类协议桥不遵循 client 载体清单：它把 core 暴露给外部生态，直接通过 `ctx.plugin(入口插件)` 挂载，不使用 fetch。

 ## 消息协议

@@ -213,7 +213,7 @@ export type ResponseValue<K> =

 | 子类 | 所在包 | doFetch | 用途 |
 |---|---|---|---|
-| `InProcessApiClient` | apiproxy 本包 | 注入的 `{ fetch }` handler | **同构点**：`new InProcessApiClient(toFetchHandler(api))` 全程不过网络但真跑 wire 序列化/zod/SSE 帧；载体测试与调用方可以在不打开端口的情况下运行这套协议，而产品 `dsh run` 直接驱动 core |
+| `InProcessApiClient` | apiproxy 本包 | 注入的 `{ fetch }` handler | **同构点**：`new InProcessApiClient(toFetchHandler(api))` 全程不过网络但真跑 wire 序列化/zod/SSE 帧；载体测试与调用方可以在不打开端口的情况下运行这套协议，而产品 `dsh --profile headless` 直接驱动 core |
 | `WebApiClient` | dsh-client-connection | `globalThis.fetch` 上行 + 每逻辑流一条同源 WebSocket 下行 | 浏览器客户端；物理边界见 [WebSocket 下行载体](2026-08-04-websocket-downlink-carrier.md) |
 | `FixtureApiClient` | dsh-client-connection | 不用（协议层覆写） | 无 server 的 UI 开发（`?fixture`）：覆写 `callUnary`/`openMux`/`openHost`/`respond` 虚方法，自己就是假 server（帧 rpcId 由它 mint，语义自洽） |
 | IPC 桥子类（假想示例——尚无此形态） | Electron 壳 | IPC 序列化往返 | 只需换 doFetch，约定/基类零改 |
@@ -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/2026-07-25-web-client-session-scope-and-provide-channel.md
-2026-07-25-web-client-session-scope-and-provide-channel.md: ff411e1b387dcd83b04f8c02d5dde3c70a815932
-2026-07-25-web-client-session-scope-and-provide-channel.zh.md: 593e194654bb3552ad3cac4534847fba11cf3694
+2026-07-25-web-client-session-scope-and-provide-channel.md: f371b93ccf6cb3ba10cbdb73baa670cbf889f393
+2026-07-25-web-client-session-scope-and-provide-channel.zh.md: 78440131c4e09c9a458009fbe5a2a34a707a481c
@@ -93,7 +93,7 @@ Slot scope is the closed set `root | session-maybe | session`:
 - `session-maybe` follows the current session with ADOPTION identity (the only behavior — there is no hold-identity-forever mode): an incarnation born session-less keeps its React instance across the arrival of the FIRST session (the blank shell adopts it — no remount, the DOM survives), and from then on behaves exactly like a strict session entry — switching to a different session remounts, and dropping back to no-session remounts into a fresh blank incarnation that will adopt again. Component-local per-session state therefore clears by construction; state that must survive a switch belongs in session-bound sources (machine, store, hooks). With no session, `sessionId`, the results of `useSession`/`useInput`, and `inputActions` may all be absent. The unkeyed root `SessionMaybeProvider` drives these updates by subscribing to the runtime's atomic `currentProvide` projection — selection moves and provider-roster changes publish through the same source, so a roster change under a stable current id republishes the mounted bundle instead of stranding entries on an obsolete hook/prop schema — while `SessionMaybeProvideInfo` uses the static key map to retain the complete hook/prop shape even with no session; the per-entry adoption bookkeeping (incarnation-counter key) lives in the renderer's `SessionMaybeEntry`.
 - `session` guarantees that `sessionId`, every hook source, and every prop exist; each strict entry's error boundary is keyed by `sessionId`, so switching sessions recreates that entry and its session store.

-`conversation` is the resident `session-maybe` shell: `ConversationRoot`, HeroShell, the Workspace picker, the root-owned scrollport and composer stack, and the overlay chain's fallback frame retain their React instances across the no-session → blank-session switch. Two strict entries fill fixed regions without reparenting that tree: `conversation.session.header` carries breadcrumb/tabs/actions above the scrollport, while `conversation.session` carries the view ring and draft mirror inside it; both share the same session-scoped chat store. The composer bar (`conversation.composer.bar`) is itself `session-maybe`: with no session it renders inert (machine faces absent, `disabled` owner prop), and the same instance — textarea included — goes live when a session appears; the remaining input slots stay strict `session` and dispatch nothing until then. The blank → engaging/active transition never rebuilds the InputBar on a phase flip.
+`conversation` is the resident `session-maybe` shell: `ConversationRoot`, HeroShell, the Workspace picker, the root-owned scrollport and composer stack, and the overlay chain's fallback frame retain their React instances across the no-session → blank-session switch. Two strict entries fill fixed regions without reparenting that tree: `conversation.session.header` carries breadcrumb/tabs/actions above the scrollport, while `conversation.session` carries the view ring and draft mirror inside it; both share the same session-scoped chat store. The composer bar (`conversation.composer.bar`) is itself `session-maybe`: with no session its machine faces and message actions are inert, while the whole dashed card opens the existing Workspace picker by pointer and its read-only textarea does the same through Enter or Space. The same instance — textarea included — goes live when a session appears; the remaining input slots stay strict `session` and dispatch nothing until then. The blank → engaging/active transition never rebuilds the InputBar on a phase flip.

 - The runtime's first built-in entry: the `'session'` hook — `useSession` itself rides the same mechanism, no special-casing.
 - Concurrent discipline: the render plane reads only from the hooks compartment (uSES consistency guarantee); props-compartment callbacks are used only in event-handler space; descriptor resolution is render-safe (idempotent caching, with prune reaping residue from abandoned renders).
@@ -132,5 +132,5 @@ Slot scope is the closed set `root | session-maybe | session`:
 - Plugins gain session context isomorphic to the host's: per-session state hangs on the actx and mounts/tears down in one piece with the scope fiber, making leaks structurally impossible; two-session isolation is structurally guaranteed by the scope filter.
 - The client object layer converges to a wire mirror: session identity, lifecycle, and capability adjudication all defer to the host entity — the input system (the next layer) always faces a session with a real Agent, and providers like slash/skill uniformly address by sessionId directly.
 - Blank-session governance takes zero dedicated mechanisms: state rides one derived bit, visibility rides the unified list projection (only the current blank shows, as `New Session`), reclamation rides lazy persistence's existing contract (evaporation on restart), and the ordinary ceiling rides same-Workspace reuse.
- The cost: the id→ctx handoff discipline and provide's Concurrent discipline are conventions rather than type-enforced, pinned by review and tests; fully disabled input while no workspace is picked is an experience cost the product surface accepts (the price of the single state axis).
+- The cost: the id→ctx handoff discipline and provide's Concurrent discipline are conventions rather than type-enforced, pinned by review and tests. The single state axis still withholds machine faces until a Session exists; the resident card routes activation to the Workspace picker during that interval ([decision](../feature/2026-08-07-workspace-picker-composer-entry.md)).
 - Known gaps: approval/question recovery across prune (TODO); model selection returns in live-mutation shape (the host `selectModel` trio is ready-made, its client consumer not yet built).
@@ -93,7 +93,7 @@ slot scope 是闭集 `root | session-maybe | session`：
 - `session-maybe` 以**收养（adoption）身份语义**跟随 current session（唯一行为——不存在「永久保持实例」模式）：空态出生的化身在**第一个** session 到来时保持 React 实例（空壳收养它——不重挂，DOM 存活）；此后行为与严格 session entry 完全一致——切到不同 session 重挂，跌回无 session 也重挂为崭新的空态化身（之后再次收养）。因此组件本地的 per-session 状态**由构造保证**随切换清零；需要活过切换的状态必须住 session 绑定的源（machine、store、hooks）。无 session 时 `sessionId`、`useSession`/`useInput` 的选择结果及 `inputActions` 均可缺省。根部无 key 的 `SessionMaybeProvider` 通过订阅 runtime 的原子 `currentProvide` 投影驱动这条更新——选择移动和提供方名册变化经同一 source 发布，current id 不变时的名册变化也会重发已挂载 bundle，而不是把 entry 困在过期的钩子/prop 形状上——`SessionMaybeProvideInfo` 靠静态键表在无 session 时仍保留完整钩子/prop 形状；逐 entry 的收养记账（化身计数 key）住在 renderer 的 `SessionMaybeEntry`。
 - `session` 保证 `sessionId`、所有钩子 source 与 props 均存在；每个严格 entry 的错误边界以 `sessionId` 为 key，切换 session 会重建该 entry 及其 session store。

-`conversation` 是 `session-maybe` 的常驻外壳：`ConversationRoot`、HeroShell、Workspace picker、root 持有的 scrollport 与 composer stack，以及 overlay chain 的 fallback 外框，在无 session → blank session 的切换中保持 React 实例。两个严格 session entry 只填入固定区域，不改变该树的父级：`conversation.session.header` 在 scrollport 上方承载 breadcrumb／tab／action，`conversation.session` 在其内部承载 view ring 与 draft mirror；二者共享同一个 session scope chat store。composer bar（`conversation.composer.bar`）本身即为 `session-maybe`：无 session 时以惰性态渲染（machine face 缺席、`disabled` owner prop），session 出现后同一实例（含 textarea）转为 live；其余输入 slot 保持严格 `session`，在此之前不分发任何条目。blank → engaging/active 的 InputBar 不因 phase 翻转而重建。
+`conversation` 是 `session-maybe` 的常驻外壳：`ConversationRoot`、HeroShell、Workspace picker、root 持有的 scrollport 与 composer stack，以及 overlay chain 的 fallback 外框，在无 session → blank session 的切换中保持 React 实例。两个严格 session entry 只填入固定区域，不改变该树的父级：`conversation.session.header` 在 scrollport 上方承载 breadcrumb／tab／action，`conversation.session` 在其内部承载 view ring 与 draft mirror；二者共享同一个 session scope chat store。composer bar（`conversation.composer.bar`）本身即为 `session-maybe`：无 session 时，其 machine face 和消息操作保持惰性，整张虚线卡片可经指针打开现有 Workspace picker，只读 textarea 也可通过 Enter 或 Space 打开。session 出现后同一实例（含 textarea）转为 live；其余输入 slot 保持严格 `session`，在此之前不分发任何条目。blank → engaging/active 的 InputBar 不因 phase 翻转而重建。

 - 运行时内建第一条：`'session'` 钩子——`useSession` 本身走同一机制，无特判。
 - Concurrent 纪律：渲染平面只从 hooks 格读（uSES 一致性保证）；props 格回调只在事件 handler 空间用；描述符解析 render-safe（幂等缓存、废弃渲染残留由 prune 收尸）。
@@ -132,5 +132,5 @@ slot scope 是闭集 `root | session-maybe | session`：
 - 插件获得与 host 同构的会话上下文：逐会话状态挂 actx、随 scope fiber 一次拆装，泄漏结构性不可能；双会话隔离由 scope filter 结构性保证。
 - client 对象层收敛为 wire 镜像：会话身份、生命周期、能力判别全部以 host 实体为准——输入体系（下一层）面对的永远是「有真 Agent 的会话」，slash/skill 等提供方一律以 sessionId 直接寻址。
 - 空会话治理零专用机制：状态靠一个派生位，可见性靠统一列表投影（仅 current blank 以 `New Session` 展示），回收靠 lazy persistence 的既有约定（重启蒸发），常规上限靠同 Workspace 复用。
- 代价：id→ctx 换乘纪律、provide 的 Concurrent 纪律都是约定而非类型强制，靠 review 与测试钉住；「未选 workspace」期间输入全禁是产品面接受的体验代价（单一状态轴换来的）。
+- 代价：id→ctx 换乘纪律、provide 的 Concurrent 纪律都是约定而非类型强制，靠 review 与测试钉住。单一状态轴仍会在 Session 存在前隐藏 machine face；这段时间内，常驻卡片会把激活操作转到 Workspace picker（[决策](../feature/2026-08-07-workspace-picker-composer-entry.md)）。
 - 已知欠账：approval/question 跨 prune 恢复（TODO）；模型选择以 live-mutation 形状回归（host `selectModel` 三件套现成，其 client 消费方尚未构建）。
@@ -1,33 +0,0 @@
-# Agent Note: Experimental and internal package group
-
-Status: implemented
-
-English | [中文](2026-07-28-experimental-plugin-package-group.zh.md)
-
-## Problem
-
-The [package hierarchy](../../../../packages/README.md) groups plugins by product role, but it cannot distinguish release packages from prototypes or internal-only packages. The team needs an obvious shared place for useful work that is not part of the official release.
-
-## Decision
-
-The subtree rules in [`packages/experimental/AGENTS.md`](../../../../packages/experimental/AGENTS.md) make `packages/experimental/<pkg>/` the required home for Cordis plugin packages whose whole public contract is experimental or internal-only. Package names remain `@deepseek-ai/dsh-<pkg>`.
-
-The group is the team's in-repository place to share engineering and product-manager prototypes: members can discover, run, review, and extend one another's work against the real plugin graph without implying product support.
-
-Official releases exclude this directory. A package enters a release only after moving to its product-role group; release packages cannot take runtime dependencies on packages here. Examples may use them, while any other runtime dependent also belongs here. Tests may use them as development dependencies.
-
-Experimental packages carry no stability, compatibility, migration, or support promise: they may change APIs, configuration, or data, or disappear without deprecation or migration. Internal-only packages may define narrower internal contracts but make no public release promise. Neither status relaxes engineering, security, documentation, lifecycle, testing, or snapshot requirements.
-
-The pending `@deepseek-ai/dsh-tui-session-changes` `/diff` viewer and `/btw` plugin are examples governed by this rule. Promotion into an official release requires explicit review of the public contract, limitations, test evidence, and a named owner accepting stable-package obligations.
-
-## Alternatives considered
-
-**Keep experimental and internal-only packages in product-role groups with README labels.** Labels are easy to miss and cannot enforce dependency boundaries.
-
-**Treat every package as experimental until the first tagged release.** This provides no durable incubation boundary.
-
-**Develop prototypes and internal packages elsewhere.** This loses the real plugin graph, examples, snapshots, and lifecycle checks needed to evaluate them.
-
-## Consequences
-
-The path makes release exclusion and dependency blast radius visible while retaining the real plugin graph for team sharing. It gives up product-role colocation and creates path churn on promotion, while the npm name remains stable. The subtree rules, repository [current-owner/current-need rule](../../../../packages/AGENTS.md), and unchanged engineering gates limit junk-drawer growth. Because official release tooling does not yet exist, contributor policy enforces the exclusion; when such tooling is added, the directory is its required exclusion boundary.
@@ -1,33 +0,0 @@
-# Agent Note: 实验性与内部专用包分组
-
-Status: implemented
-
-[English](2026-07-28-experimental-plugin-package-group.md) | 中文
-
-## 问题
-
-[包层级结构](../../../../packages/README.md)按产品角色对插件分组，但无法区分发布包、原型和内部专用包。团队需要一个明确的共享位置，存放不属于官方发布版本的有价值成果。
-
-## 决策
-
-[`packages/experimental/AGENTS.md`](../../../../packages/experimental/AGENTS.md) 中的子树规则要求所有公开约定整体处于实验状态或仅限内部使用的 Cordis 插件包位于 `packages/experimental/<pkg>/`。包名仍为 `@deepseek-ai/dsh-<pkg>`。
-
-该分组供团队在仓库内共享工程人员和产品经理制作的原型：成员可以基于真实插件图发现、运行、评审并扩展彼此的原型，但这不代表产品会提供支持。
-
-官方发布版本不包含此目录。包只有移入对应的产品角色分组后才会纳入发布版本；发布包不得在运行时依赖此处的包。示例可以使用这些包；其他任何运行时依赖方也必须位于此处。测试可以将它们用作开发依赖。
-
-实验性包不提供稳定性、兼容性、迁移或支持保证：其 API、配置或数据可以变更，包也可以移除，均不提供弃用期或迁移路径。内部专用包可以定义范围更窄的内部约定，但不作公开发布承诺。无论哪种状态，都不降低仓库对工程、安全、文档、生命周期、测试或快照的要求。
-
-尚待完成的 `@deepseek-ai/dsh-tui-session-changes` `/diff` 查看器和 `/btw` 插件都受这项规则约束。将包提升为稳定包并纳入官方发布版本，需要明确评审其公开约定、限制和测试证据，并指定一名愿意承担稳定包义务的负责人。
-
-## 考虑过的替代方案
-
-**将实验性和内部专用包留在产品角色分组中，并用 README 标注。** 标注容易被忽略，也无法强制执行依赖边界。
-
-**首个带标签的版本发布前，将所有包都视为实验性。** 这无法提供持久的孵化边界。
-
-**在其他位置开发原型和内部专用包。** 这会失去评估它们所需的真实插件图、示例、快照和生命周期检查。
-
-## 后果
-
-该路径明确标示不纳入发布版本的包及其依赖影响范围，同时保留供团队共享成果的真实插件图。代价是这些包无法与同产品角色的包共置，提升并纳入发布版本时还会产生路径变动，但 npm 包名保持稳定。子树规则、仓库已有的[「必须有当前负责人和实际需求」规则](../../../../packages/AGENTS.md)以及保持不变的工程门禁，可限制该分组无序膨胀。由于官方发布工具尚不存在，目前由贡献者政策执行这项排除规则；添加发布工具后，必须以该目录为排除边界。
@@ -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/2026-07-29-dsh-source-launch-tsx-esm.md
-2026-07-29-dsh-source-launch-tsx-esm.md: ed22e51d59a25db130b3760ce484c116bade4348
-2026-07-29-dsh-source-launch-tsx-esm.zh.md: bdd549092eb30f7749c8f7561068daafe3548b28
+2026-07-29-dsh-source-launch-tsx-esm.md: b2428602a780f2880f0f803ba59b16a76b39790e
+2026-07-29-dsh-source-launch-tsx-esm.zh.md: 866bc8886725788e6e619f9507f5c8ccab63f82d
@@ -14,11 +14,11 @@ Startup latency also mattered: the off-thread `module.register()` hooks worker s

 ## Decision

-The `dsh` TUI, Web, and headless source launches run `node --import tsx/esm`: tsx's ESM-only hook owns both TypeScript transformation and tsconfig `paths` projection. `bin/dsh`, the root `dsh`/`demo:tui`/`demo:web` scripts, and the Code Mode TUI overlay use the same vector; `bin/dsh` references the hook and tsconfig by absolute checkout paths (bare `tsx/esm` does not resolve from an arbitrary cwd) and pins `TSX_TSCONFIG_PATH` to the root tsconfig. The CJS hook stays off because the CLI source graph is ESM-only; measured TUI time-to-banner is ~0.7s versus ~1.1s under the full tsx default and ~0.75s under the removed native chain.
+The `dsh` TUI, Web, and headless source launches run `node --import tsx/esm`: tsx's ESM-only hook owns both TypeScript transformation and tsconfig `paths` projection. The root `dsh` script completes the repository build, then uses that vector from the repository root. The CJS hook stays off because the CLI source graph is ESM-only; measured runtime launch to the TUI banner is ~0.7s versus ~1.1s under the full tsx default and ~0.75s under the removed native chain.

 `scripts/tspath-loader.ts` and `apps/cli/src/tsconfig-paths-loader.ts` are deleted. With them went the loader's runtime rule of mapping a workspace import only for declared runtime dependencies — tsx applies the `paths` map unconditionally. Declaration completeness now rests on the static gates alone: `verify-cordis-config` for configured bare plugins, and workspace constraints for manifests. (That runtime rule found real bugs: `dsh-plan-mode` and `dsh-tool-tasks` imported `@deepseek-ai/dsh-llm` while declaring it only in devDependencies; since fixed.)

-The node-compat CI matrix (Node 22.19 and 26) gains `dsh-source-launch-smoke` (`apps/cli/tests/source-launch.compat.spec.ts`): a keyless piped-stdio launch of the exact production vector asserting the non-zero-exit TTY refusal. Any future Node change to module hooks or TypeScript handling turns this gate red instead of breaking developers' `pnpm dsh`.
+The node-compat CI matrix (Node 22.19 and 26) gains `dsh-source-launch-smoke` (`apps/cli/tests/source-launch.compat.spec.ts`): a keyless piped-stdio launch of the exact production runtime vector asserting the non-zero-exit TTY refusal. Any future Node change to module hooks or TypeScript handling turns this gate red instead of breaking developers' `pnpm dsh`.

 ## Alternatives considered

@@ -35,4 +35,4 @@ The node-compat CI matrix (Node 22.19 and 26) gains `dsh-source-launch-smoke` (`
 - One launch vector across the whole engines range, including future Node lines that change native TypeScript support; the smoke gate enforces it per matrix line.
 - TypeScript transformation is delegated to tsx/esbuild again, reversing the prior note's goal of proving Node-native transformation; that goal is unreachable while vendored sources use non-erasable syntax and Node ships no transform mode.
 - The runtime declared-dependency enforcement in source launches is gone; undeclared workspace imports now surface only through static gates or built-mode resolution failures.
- Startup improves ~0.4s over the full tsx default (`demo:headless` now aliases the same `dsh run` source launch; ACP keeps `--import tsx` because its graph was not audited for CJS-hook dependence and its launch latency is not on the interactive path).
+- Runtime launch improves ~0.4s over the full tsx default; ACP keeps `--import tsx` because its graph was not audited for CJS-hook dependence and its launch latency is not on the interactive path.
@@ -14,11 +14,11 @@ Status: implemented

 ## 决策

-`dsh` 的 TUI、Web 与无头源码启动运行 `node --import tsx/esm`：由 tsx 的 ESM-only 钩子同时负责 TypeScript 转换与 tsconfig `paths` 投影。`bin/dsh`、根目录的 `dsh`/`demo:tui`/`demo:web` 脚本以及 Code Mode TUI overlay 使用同一向量；`bin/dsh` 以 checkout 的绝对路径引用钩子与 tsconfig（裸的 `tsx/esm` 无法从任意 cwd 解析），并将 `TSX_TSCONFIG_PATH` 固定到根 tsconfig。CJS 钩子保持关闭，因为 CLI 源码图是纯 ESM；实测 TUI 到 banner 约 0.7s，对比完整 tsx 默认形态约 1.1s、已移除的原生链约 0.75s。
+`dsh` 的 TUI、Web 与无头源码启动运行 `node --import tsx/esm`：由 tsx 的 ESM-only 钩子同时负责 TypeScript 转换与 tsconfig `paths` 投影。根目录的 `dsh` 脚本先完成仓库构建，然后从仓库根目录使用同一启动方式。CJS 钩子保持关闭，因为 CLI（命令行界面）源码图是纯 ESM；实测运行时启动至 TUI banner 耗时约 0.7s，对比完整 tsx 默认形态约 1.1s、已移除的原生链约 0.75s。

 `scripts/tspath-loader.ts` 与 `apps/cli/src/tsconfig-paths-loader.ts` 已删除。随之消失的还有该 loader「仅为已声明运行时依赖映射 workspace import」的运行时规则——tsx 无条件应用 `paths` 映射。声明完整性现在仅由静态门禁保障：配置的裸插件走 `verify-cordis-config`，manifest（元数据清单）走 workspace constraints。（该运行时规则确实发现过真实缺陷：`dsh-plan-mode` 与 `dsh-tool-tasks` 导入 `@deepseek-ai/dsh-llm` 却只声明在 devDependencies；后已修复。）

-node-compat CI 矩阵（Node 22.19 与 26）新增 `dsh-source-launch-smoke`（`apps/cli/tests/source-launch.compat.spec.ts`）：以精确的生产启动向量做 keyless 管道 stdio 启动，断言非零退出的 TTY 拒绝。未来 Node 对模块钩子或 TypeScript 处理的任何改动都会让该门禁变红，而不是破坏开发者的 `pnpm dsh`。
+node-compat CI 矩阵（Node 22.19 与 26）新增 `dsh-source-launch-smoke`（`apps/cli/tests/source-launch.compat.spec.ts`）：以精确的生产运行时启动向量做 keyless 管道 stdio 启动，断言非零退出的 TTY 拒绝。未来 Node 对模块钩子或 TypeScript 处理的任何改动都会让该门禁变红，而不是破坏开发者的 `pnpm dsh`。

 ## 备选方案

@@ -35,4 +35,4 @@ node-compat CI 矩阵（Node 22.19 与 26）新增 `dsh-source-launch-smoke`（`
 - 整个 engines 范围（包括未来改变原生 TypeScript 支持的 Node 版本线）只有一个启动向量；冒烟门禁按矩阵行强制执行。
 - TypeScript 转换重新委托给 tsx/esbuild，逆转了前一篇 Agent Note「证明 Node 原生转换可用」的目标；在 vendor 源码使用不可擦除语法且 Node 不再提供 transform 模式的情况下，该目标不可达。
 - 源码启动中的运行时依赖声明强制不复存在；未声明的 workspace import 现在只能通过静态门禁或构建模式的解析失败暴露。
- 启动相比完整 tsx 默认形态快约 0.4s（`demo:headless` 现为同一条 `dsh run` 源码启动命令的别名；ACP 保留 `--import tsx`，因为它的依赖图尚未就 CJS 钩子依赖性做审计，且其启动延迟不在交互路径上）。
+- 运行时启动相比完整 tsx 默认形态快约 0.4s；ACP（Agent Client Protocol）保留 `--import tsx`，因为它的依赖图尚未就 CJS 钩子依赖性做审计，且其启动延迟不在交互路径上。
@@ -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/2026-07-29-package-regrouping.md
-2026-07-29-package-regrouping.md: 3c37bce05bacd6af800a76ac93fb691b896a6772
-2026-07-29-package-regrouping.zh.md: 68903ff1fad6a975c4445fe8971c8fe0dd40117f
+2026-07-29-package-regrouping.md: 30fc45a122263350b4a2ad1998850f631c20f9b8
+2026-07-29-package-regrouping.zh.md: a3a9a11ec71b7f894dcea7c733eb39a80b71ac50
@@ -58,7 +58,7 @@ The moves landed as pure `git mv` moves, so rename detection carries the history

 A group move did not touch: npm names, imports, `cordis.yml` configs, snapshot fixtures, the `pnpm-workspace.yaml`/`tsdown` globs (both `packages/*/*`), or the Python runtime manifest — all reference packages by npm name.

-`client/` and `host/` were out of scope and are unchanged. The `experimental/` group proposal (PR #844) is orthogonal — a release-boundary container, not a clustering decision.
+`client/` and `host/` were out of scope and are unchanged.

 ## Alternatives considered

@@ -58,7 +58,7 @@ Status: implemented

 组移动未触及：npm 包名、import、`cordis.yml` 配置、快照 fixture（测试前置数据）、`pnpm-workspace.yaml` 与 `tsdown` 的 glob（都是 `packages/*/*`），以及 Python 运行时 manifest（元数据清单）——它们全部按 npm 包名引用包。

-`client/` 与 `host/` 不在本次范围内，保持不变。`experimental/` 组提案（PR #844）与本案正交：它是发布边界容器，不是聚类决策。
+`client/` 与 `host/` 不在本次范围内，保持不变。

 ## Alternatives considered

@@ -1,47 +0,0 @@
-# Agent Note: Package-manager-native repository cache
-
-Status: implemented
-
-English | [中文](2026-07-30-package-manager-native-repository-cache.zh.md)
-
-## Problem
-
-A standalone Harness app cannot rely on a developer-owned SDK project to declare and install repository dependencies. Loading a configured GitHub repository therefore needs a persistent fetch, preparation, and cache boundary, but implementing Git transport, hosted-source syntax, package preparation, and a content store inside DSH would duplicate a package manager. Requiring a separately installed package manager would make a config-only feature depend on host setup.
-
-The cache also needs an update identity. A mutable branch name cannot both remain permanently cached and reflect later commits without an independent refresh protocol.
-
-## Decision
-
-Vendored `@cordisjs/plugin-loader/repository` exports `RepositoryCache`, a generic Node-only package helper with no DSH plugin-format knowledge. Keeping it on a subpath prevents browser consumers of the Loader's main entry from traversing Node filesystem and child-process imports. The caller supplies a package-manager-native source specifier and a cache root. DSH-specific callers own accepted source syntax, path selection, and the cache-root location; the [SDK project dependency workflow](../../proposed/feature/2026-07-17-sdk-follow-up-capabilities.md#external-cordis-plugin-installation) remains a separate path owned by the developer project's selected package manager.
-
-The Loader carries an exact runtime dependency on `pnpm@11.7.0` and invokes that package's JavaScript entry with the current Node executable. It never discovers a global executable or delegates through Corepack. Each cache miss creates an isolated project with one dependency named `repository`; pnpm owns Git/GitHub resolution, fetching, its content-addressed store, dependency installation, and lifecycle scripts in the repository's dependency graph.
-
-The isolated workspace sets `dangerouslyAllowAllBuilds: true`. A configured repository and its dependency graph are trusted executable code: lifecycle scripts may run before DSH reads any declared assets. The child receives ordinary host process state needed by Git and pnpm, but ambient credential-shaped (`KEY`, `PASSWORD`, `SECRET`, `TOKEN`) variables are removed. No OAuth, token forwarding, or private-repository authentication contract is added.
-
-The SHA-256 of the exact specifier names the cache entry. Concurrent same-process requests share one task. Installation occurs in a sibling temporary directory; only a successful install with a package directory and marker is atomically renamed into the final key. Failed staging is removed, and a competing process's already-published valid entry wins. A later process validates the marker and package directory before returning the stable `node_modules/repository` path.
-
-An identical specifier permanently reuses its published entry. The caller changes the ref or another part of the specifier to request a new generation; the cache does not poll remotes, reinterpret mutable refs, expire entries, or garbage-collect old generations.
-
-## Alternatives considered
-
-**Implement GitHub download, archive extraction, preparation, and caching directly.** Rejected under the [dependency policy](../process/2026-07-26-dependencies-over-hand-rolling.md): pnpm already owns hosted Git syntax, Git execution, lifecycle policy, and a shared content store. A second resolver would add more code while still needing package semantics.
-
-**Require `pnpm` on `PATH` or invoke Corepack.** Rejected because changing one app config must be sufficient on every supported installation. Pinning and shipping the CLI also makes the preparation policy reviewable and independent of the host's package-manager version.
-
-**Resolve a branch or tag again on every startup.** Rejected because it turns startup into a network refresh, changes code without a config diff, and makes rollback depend on remote state. Explicit ref changes preserve auditability even when a user deliberately chooses a mutable ref.
-
-**Disable repository lifecycle scripts.** Rejected because common plugin repositories need a declarative `prepare` step to validate and package their plugin subdirectory. The trust boundary is explicit configuration of executable source, not an incomplete illusion that only static files can run.
-
-**Introduce a Cordis repository service.** Rejected because cache lookup has no runtime contribution registry or provider variation. A small helper lets the later host own Cordis lifecycle and HMR without adding a service contract prematurely.
-
-## Consequences
-
- Standalone apps carry pnpm's approximately 18.6 MB unpacked runtime instead of requiring a global tool or owning a Git/package implementation.
- A repository author may use ordinary package preparation, and a malicious configured repository or dependency can execute code with the scrubbed child environment and the user's filesystem authority.
- Exact specifiers make startup deterministic after the first successful install; changing cached code requires a config/ref change.
- Failed installs leave no published cache entry and may be retried. Published corruption fails loud instead of silently reinstalling under the same identity.
- Cache generations consume disk until a future explicit cache-management policy removes them.
-
-## Testing
-
-`packages/boot/app-boot/tests/repository-cache.spec.ts` covers same-process single-flight, cross-instance cache reuse, exact-specifier separation, failed-stage cleanup and retry, and boundary validation. Its real local-Git case invokes the bundled pnpm, runs the fixture repository's `prepare` script, and reads the prepared file from the installed cache entry without network access.
@@ -1,47 +0,0 @@
-# Agent Note: 包管理器原生仓库缓存
-
-Status: implemented
-
-[English](2026-07-30-package-manager-native-repository-cache.md) | 中文
-
-## 问题
-
-独立运行的 Harness 应用不能依赖开发者自有的 SDK 工程来声明并安装仓库依赖。因此，加载配置中的 GitHub 仓库需要一道持久的获取、准备与缓存边界；但如果在 DSH 内实现 Git 传输、托管来源语法、包准备流程和内容存储，就会重复实现包管理器。若要求用户另行安装包管理器，则只需修改配置即可使用的功能还会依赖宿主环境的额外配置。
-
-缓存还需要明确更新标识。若没有独立的刷新协议，可变分支名无法既永久缓存，又反映后续 commit。
-
-## 决策
-
-vendor 中的 `@cordisjs/plugin-loader/repository` 导出 `RepositoryCache`：一个不包含 DSH 插件格式知识、仅限 Node 使用的通用包辅助工具。把它保留在子路径上，可以避免 Loader 主入口的浏览器消费方在解析依赖时遍历到 Node 文件系统和子进程 import。调用方提供包管理器原生的来源 specifier 和缓存根目录。DSH 专属调用方负责规定可接受的来源语法、路径选择与缓存根目录位置；[SDK 工程依赖工作流](../../proposed/feature/2026-07-17-sdk-follow-up-capabilities.md#external-cordis-plugin-installation)仍是另一条路径，由开发者工程选定的包管理器负责。
-
-Loader 将 `pnpm@11.7.0` 作为固定版本的运行时依赖，并使用当前 Node 可执行文件调用该包的 JavaScript 入口。它绝不探测全局可执行文件，也不经 Corepack 调用。每次缓存未命中都会创建一个隔离工程，其中只有一个名为 `repository` 的依赖；Git 与 GitHub 来源的解析和获取、pnpm 自身的内容寻址 store、依赖安装，以及仓库依赖图中的生命周期脚本均由 pnpm 负责。
-
-隔离工作区设置 `dangerouslyAllowAllBuilds: true`。用户配置的仓库及其依赖图都属于受信任的可执行代码：DSH 读取任何已声明资产之前，生命周期脚本就可能运行。子进程会收到 Git 与 pnpm 所需的常规宿主进程状态，但会移除环境中名称形似凭据（`KEY`、`PASSWORD`、`SECRET`、`TOKEN`）的变量。该机制不新增 OAuth、token 转发或私有仓库认证约定。
-
-缓存项以精确 specifier 的 SHA-256 命名。同一进程内针对相同 specifier 的并发请求共享一项任务。安装在同级临时目录中进行；只有安装成功且存在包目录和标记时，系统才会把暂存目录原子重命名为最终键对应的目录。失败的暂存目录会被删除；如果另一进程已发布有效项，则以该项为准。后续进程会先校验标记与包目录，再返回稳定的 `node_modules/repository` 路径。
-
-相同的 specifier 会永久复用已发布项。调用方通过修改 ref 或 specifier 的其他部分来请求新的缓存代次；缓存不会轮询远端、重新解释可变 ref、让条目过期，也不会垃圾回收旧代次。
-
-## 曾考虑的替代方案
-
-**直接实现 GitHub 下载、归档解压、准备与缓存。** 根据[依赖政策](../process/2026-07-26-dependencies-over-hand-rolling.md)不予采纳：pnpm 已负责托管 Git 语法、Git 执行、生命周期政策和共享内容存储。第二套解析器会增加更多代码，却仍需实现包语义。
-
-**要求 `pnpm` 位于 `PATH` 上，或调用 Corepack。** 不予采纳：在每种受支持的安装形态中，只修改一份应用配置就必须足以启用该功能。固定并随应用分发 CLI（命令行界面）还能使准备政策可供评审，并与宿主的包管理器版本无关。
-
-**每次启动都重新解析分支或 tag。** 不予采纳：这会把启动变成网络刷新，在配置 diff 未变化时更改代码，并让回滚依赖远端状态。即使用户有意选择可变 ref，显式修改 ref 仍能保持可审计性。
-
-**禁用仓库生命周期脚本。** 不予采纳：常见插件仓库需要声明式 `prepare` 步骤来校验并打包插件子目录。信任边界是显式配置可执行来源，而不是营造一种不完整的假象，仿佛只有静态文件能够运行。
-
-**引入 Cordis 仓库服务。** 不予采纳：缓存查找没有运行时贡献注册表，也不存在提供方变体。小型 helper 让后续宿主负责 Cordis 生命周期与 HMR（热模块替换），无需过早新增服务约定。
-
-## 后果
-
- 独立应用随附 pnpm 约 18.6 MB 的解压后运行时，不要求全局工具，也无需自行实现 Git 与包处理。
- 仓库作者可以使用常规包准备流程；恶意的已配置仓库或依赖可以在经过上述清理的子进程环境中，以用户的文件系统权限执行代码。
- 精确 specifier 使首次安装成功后的启动具有确定性；更改缓存代码必须修改配置或 ref。
- 安装失败不会留下已发布缓存项，可以再次重试。已发布缓存损坏时会明确报错，而不会在同一标识下静默重装。
- 缓存代次会持续占用磁盘，直到未来有明确的缓存管理政策将其移除。
-
-## 测试
-
-`packages/boot/app-boot/tests/repository-cache.spec.ts` 覆盖同进程 single-flight、跨实例缓存复用、精确 specifier 隔离、失败暂存清理与重试，以及边界校验。其真实本地 Git 用例会调用随附的 pnpm，运行 fixture（测试前置数据）仓库的 `prepare` 脚本，并在不访问网络的情况下，从已安装缓存项中读取准备后的文件。
@@ -1,49 +0,0 @@
-# Agent Note: Static repository Plugin format
-
-Status: implemented
-
-English | [中文](2026-07-30-static-repository-plugin-format.zh.md)
-
-## Problem
-
-A repository that already contains reusable skills or an MCP server declaration should be usable by standalone Harness applications without becoming a Harness SDK project or rewriting its existing layout. Popular repositories must be able to add one `.dsh-plugin` directory while keeping their current skills and `.mcp.json` elsewhere in the tree. These portable static contributions still need to reuse the existing skill and MCP lifecycle owners when the same trusted package also carries native Cordis code.
-
-The [package-manager-native repository cache](2026-07-30-package-manager-native-repository-cache.md) prepares an exact package source but intentionally knows nothing about DSH formats. This layer therefore needs a package-manager-compatible authoring format, a deterministic prepared artifact, and a Cordis composition that stays transactional under Loader disposal and replacement.
-
-## Decision
-
-`@deepseek-ai/dsh-repository-plugin` owns the static contribution subformat inside a `.dsh-plugin` package: skill roots and one common `.mcp.json`. Its package metadata uses `package.json#dsh.skills` for relative skill-root paths and `package.json#dsh.mcpServers` for the relative MCP document path. Each path may leave `.dsh-plugin` to reuse repository content but must remain beneath the directory containing that `.dsh-plugin`; a nested selectable Plugin therefore owns the adjacent subtree above its package without gaining access to unrelated host paths. The package may additionally declare the explicit code entry owned by the [trusted repository package decision](2026-08-08-trusted-repository-package-code.md), and at least one code or static contribution is required.
-
-The `.dsh-plugin` package declares the published `@deepseek-ai/dsh-repository-plugin` package as a development dependency and a non-empty `scripts.prepack` that invokes its `dsh-plugin-prepare` executable. During Git installation, pnpm installs that dependency from the selected package's own manifest; `prepack` runs after dependency installation and before pnpm packs a selected subdirectory, including a Plugin nested inside another package-manager workspace. The package may build its code first. The helper validates metadata and source types, strictly parses `.mcp.json`, copies static assets into `dsh-plugin-assets`, and writes `dsh-plugin.mjs`; the source loader revalidates the installed package's helper-bearing lifecycle metadata before importing that wrapper. A static-only package still receives an import-free wrapper containing its normalized manifest, service-derived `inject` list, and delegation to the `dsh-repository-plugin` Loader builtin. The dependency and workspace-isolation rationale is in the [Git source preparation repair](../bug-fix/2026-08-08-npm-backed-git-repository-plugin-preparation.md).
-
-Loading the DSH package registers that builtin as an effect. A generated wrapper mounts the builtin as its child with `import.meta.url`, so all contributions belong to the wrapper fiber and disappear on Loader removal or rollback. The builtin revalidates the prepared manifest and path containment before reading assets. It composes the existing implementations rather than registering skills or MCP tools itself.
-
-Each prepared skill set mounts `dsh-skill-local` with a unique `repository:<package-name>` provider name, only the copied custom roots, and watching disabled. `dsh-skill-local` therefore gains two general configuration fields: `providerName` and `includeDefaultRoots`. Their defaults preserve its existing single local provider; repository instances set a distinct name and exclude project/user roots so multiple instances neither collide nor duplicate host-local discovery.
-
-Each `.mcp.json` server becomes one existing `dsh-mcp-client` child. The adapter accepts the common root `{ "mcpServers": ... }`; stdio definitions allow only optional `type: "stdio"`, `command`, `args`, and `env`, while HTTP definitions allow only `type: "http"`, `url`, and `headers`. Exact `${NAME}` process-environment references expand at runtime, after cache preparation; missing names fail Plugin load. HTTP maps to the client's Streamable HTTP transport, and stdio uses the prepared package directory as `cwd`. The existing client alone owns connection attempts, failure logging, remote tool synchronization, tool calls, and disconnects. Repository instances enable strict startup, so an initial connection, discovery, or tool-registration failure rejects the repository Loader generation; non-strict standalone clients retain the logged successful-plugin/no-tools behavior.
-
-Unknown MCP fields reject. This intentionally excludes OAuth, `auth` objects, `CLAUDE_PLUGIN_ROOT`, and a broader Claude compatibility contract. Commands, hooks, agents, rules, and other foreign manifest conventions are not inferred from static repository layout; DSH-native behavior uses the explicit trusted Cordis entry. Repository subdirectory selection and GitHub source configuration belong to the [standalone app integration](../feature/2026-07-30-config-only-repository-plugins.md), not this static adapter.
-
-## Alternatives considered
-
-**Discover an entry from `main`, `exports`, or repository layout.** Rejected because static assets do not imply that a package's ordinary entry is a Cordis Plugin. Trusted code loading is explicit through `dsh.entry` and remains outside this static adapter's ownership.
-
-**Teach generated wrappers to implement skills and MCP directly.** Rejected because copied runtime code would drift from `dsh-skill-local` and `dsh-mcp-client`, especially their provider invalidation, tool synchronization, failure, and teardown contracts.
-
-**Import Harness packages from each generated wrapper.** Rejected because repository packages should not resolve or version the application's internal dependency graph. A Loader builtin supplies one app-owned implementation and keeps generated wrappers import-free.
-
-**Watch prepared repository assets.** Rejected because an exact repository cache generation is immutable. Ref, subdirectory, or configuration changes select a new generation; a second watcher would create an unowned refresh identity.
-
-**Make every MCP connect failure a Loader update failure.** Rejected because optional standalone MCP clients deliberately contain startup failures and expose no tools. The MCP client instead owns an explicit strict-startup option, which repository adapters enable for their declared servers.
-
-## Consequences
-
- Existing skill/MCP repositories can add a small `.dsh-plugin/package.json` without relocating their assets or adopting an SDK project.
- Prepared static output is deterministic glue, while an optional `dsh.entry` and the configured repository lifecycle remain trusted executable package-manager input rather than a sandbox.
- Multiple repository Plugins coexist through provider names and ordinary MCP server-name uniqueness; duplicate names fail through their existing registries and participate in Loader rollback.
- Cached source edits do not appear live. Another exact source/ref/path/config selection is required.
- Adding another portable static contribution kind requires an explicit format and DSH-owned runtime consumer; DSH-native behavior uses the separate explicit code entry.
-
-## Testing
-
-Focused tests prepare skills and MCP metadata, prove a static-only wrapper contains no imports, reject Work IQ-style OAuth fields, map Expo-style HTTP and DataJunction-style stdio plus environment values, and exercise missing variables. A real Loader test mounts a generated wrapper through the registered builtin, reads its skill through `ctx.skills`, removes the Loader entry, and observes provider cleanup. The CI built-entry acceptance invokes `dsh run` with a GitHub source pinned to the pull request head and observes the copied skill alongside the trusted code and MCP proofs owned by the superseding decision.
@@ -1,49 +0,0 @@
-# Agent Note: 静态 repository Plugin 格式
-
-状态：已实现
-
-[English](2026-07-30-static-repository-plugin-format.md) | 中文
-
-## 问题
-
-一个已经包含可复用 skills 或 MCP server 声明的仓库，应当能被独立 Harness 应用使用，而不必先变成 Harness SDK 项目，也不应被迫改写现有布局。常见仓库只需新增一个 `.dsh-plugin` 目录，同时仍可把原有 skills 与 `.mcp.json` 放在仓库其他位置。当同一个受信任包还携带原生 Cordis 代码时，这些可移植静态贡献仍需复用现有的 skill 与 MCP 生命周期所有者。
-
-[Package-manager-native repository cache](2026-07-30-package-manager-native-repository-cache.md) 会准备一个精确 package source，但有意不了解任何 DSH 格式。因此本层需要一种兼容 package manager 的创作格式、确定性的已准备产物，以及在 Loader dispose 和替换期间仍保持事务性的 Cordis 组合。
-
-## 决策
-
-`@deepseek-ai/dsh-repository-plugin` 负责 `.dsh-plugin` 包内的静态贡献子格式：skill 根和一个通用 `.mcp.json`。其包元数据使用 `package.json#dsh.skills` 声明相对 skill 根路径，使用 `package.json#dsh.mcpServers` 声明相对 MCP 文档路径。每条路径都可以离开 `.dsh-plugin` 以复用仓库内容，但必须留在包含该 `.dsh-plugin` 的目录之下；因此，一个嵌套且可选择的插件可以拥有其包上方相邻的子树，却不能访问无关宿主路径。该包还可以声明由[受信任 repository 包决策](2026-08-08-trusted-repository-package-code.md)负责的显式代码入口，并且至少需要一种代码或静态贡献。
-
-`.dsh-plugin` 包将已发布的 `@deepseek-ai/dsh-repository-plugin` 包声明为开发依赖，并声明非空 `scripts.prepack` 来调用其 `dsh-plugin-prepare` 可执行文件。在 Git 安装期间，pnpm 会按所选包自身的 manifest（元数据清单）安装该依赖；`prepack` 会在依赖安装后、pnpm 打包选定子目录前运行，即使插件嵌套在另一个包管理器工作区内也不例外。包可以先构建其代码。该辅助程序会校验元数据与源码类型，严格解析 `.mcp.json`，把静态资源复制到 `dsh-plugin-assets`，并写入 `dsh-plugin.mjs`；源码 loader 会在导入该包装层前重新校验已安装包的生命周期元数据是否包含辅助命令。仅含静态贡献的包仍会获得无 import 包装层，其中包含规范化 manifest、由服务派生的 `inject` 列表，以及对 `dsh-repository-plugin` Loader builtin 的委托。依赖与 workspace 隔离的设计依据见[Git 源准备修复](../bug-fix/2026-08-08-npm-backed-git-repository-plugin-preparation.md)。
-
-加载 DSH package 会以 effect 方式注册该 builtin。生成的包装模块使用 `import.meta.url` 把 builtin 挂载为自己的子级，因此所有贡献都归属于包装 fiber，并在 Loader 移除或回滚时消失。Builtin 会在读取资源前重新校验已准备 manifest 与路径包含关系。它只组合现有实现，而不自行注册 skills 或 MCP 工具。
-
-每份已准备 skill 集合都会挂载 `dsh-skill-local`，使用唯一的 `repository:<package-name>` 提供方名称、仅包含复制后的自定义根，并禁用监视。因此 `dsh-skill-local` 新增两个通用配置字段：`providerName` 和 `includeDefaultRoots`。默认值保持原有单一本地提供方行为；repository 实例设置不同名称并排除项目／用户根，使多个实例既不冲突，也不会重复宿主本地发现。
-
-`.mcp.json` 中的每个 server 都变成一个现有 `dsh-mcp-client` 子级。适配层接受通用根对象 `{ "mcpServers": ... }`；stdio 定义只允许可选的 `type: "stdio"`、`command`、`args` 与 `env`，HTTP 定义只允许 `type: "http"`、`url` 与 `headers`。严格的 `${NAME}` 进程环境变量引用在运行时、cache 准备之后展开；缺失变量会使 Plugin 加载失败。HTTP 映射到 client 的 Streamable HTTP transport，stdio 使用已准备 package 目录作为 `cwd`。只有现有 client 负责连接尝试、失败日志、远端工具同步、工具调用和断开。Repository 实例会启用严格启动，因此初始连接、发现或工具注册失败会拒绝 repository Loader generation；非严格的独立 client 则保留“记录日志、Plugin 成功但不注册工具”的行为。
-
-未知 MCP 字段会被拒绝。这里有意排除 OAuth、`auth` 对象、`CLAUDE_PLUGIN_ROOT` 和更广泛的 Claude 兼容约定。命令、hook、agent（智能体）、规则和其他外来 manifest 约定不会从静态 repository 布局中推断出来；DSH 原生行为使用显式的受信任 Cordis 入口。Repository 子目录选择与 GitHub 源配置属于[独立应用集成](../feature/2026-07-30-config-only-repository-plugins.md)，而不是本静态适配器。
-
-## 考虑过的替代方案
-
-**从 `main`、`exports` 或 repository 布局中发现入口。** 拒绝，因为静态资源并不表示包的普通入口就是 Cordis 插件。受信任代码通过 `dsh.entry` 显式加载，不属于该静态适配器的职责。
-
-**让生成包装模块直接实现 skills 和 MCP。** 拒绝，因为复制的运行时代码会与 `dsh-skill-local` 和 `dsh-mcp-client` 漂移，尤其是提供方失效、工具同步、失败和 teardown 约定。
-
-**让每个生成包装模块 import Harness package。** 拒绝，因为 repository package 不应解析或锁定应用的内部依赖图。Loader builtin 提供一份由 app 所有的实现，并让生成包装模块保持无 import。
-
-**监视已准备 repository 资源。** 拒绝，因为一个精确 repository cache generation 是不可变的。Ref、子目录或配置变化会选择新 generation；第二套 watcher 会创造一套没有所有者的刷新身份。
-
-**把每次 MCP 连接失败都当作 Loader 更新失败。** 拒绝，因为可选的独立 MCP client 会有意收束启动失败，并且不暴露工具。MCP client 改为自行提供显式的严格启动选项，由 repository 适配器为其声明的 server 启用。
-
-## 后果
-
- 现有 skill／MCP 仓库可以新增一个很小的 `.dsh-plugin/package.json`，无需移动资源或采用 SDK 项目。
- 已准备的静态输出是确定性胶水；可选的 `dsh.entry` 和已配置的 repository 生命周期仍是受信任的可执行包管理器输入，而非沙箱。
- 多个 repository Plugin 通过提供方名称和普通 MCP server-name 唯一性共存；重复名称经现有 registry 失败，并参与 Loader 回滚。
- Cache 内的源码编辑不会实时出现；必须选择另一个精确 source／ref／path／config。
- 新增可移植静态贡献类型必须提供显式格式和 DSH 自有运行时消费方；DSH 原生行为使用独立的显式代码入口。
-
-## 测试
-
-聚焦测试会准备 skill 与 MCP 元数据，证明仅含静态贡献的包装模块不含 import，拒绝 Work IQ 风格的 OAuth 字段，映射 Expo 风格 HTTP 与 DataJunction 风格 stdio 及环境变量，并覆盖缺失变量。真实 Loader 测试通过已注册 builtin 挂载生成包装模块，经 `ctx.skills` 读取其 skill，移除 Loader 条目并观察提供方清理。CI 构建入口验收会使用锁定到 PR（Pull Request）head 的 GitHub 源调用 `dsh run`，并观察已复制的 skill，以及由取代本决策的新决策所负责的受信任代码与 MCP 验证证据。
@@ -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/2026-08-03-per-session-agent-presets.md
-2026-08-03-per-session-agent-presets.md: c39117ab0de001650a95f98ccf3e42f3a5034c92
-2026-08-03-per-session-agent-presets.zh.md: 5e98a2865013a355c134317ded8a4f2ddaccf42c
+2026-08-03-per-session-agent-presets.md: 5a82f0220058c10892b819a83499c817aa9be6ad
+2026-08-03-per-session-agent-presets.zh.md: 0fb3330014eb703c854f70c9e5ed552256651c31
@@ -18,7 +18,7 @@ Composition splits into two planes, decided by what must be shared rather than b

 | Plane | Instances | Contents |
 |---|---|---|
-| Host | one | The registries themselves (`tools`, `systemPrompt`, `agents`, `agent-loop`, `sessions`), cross-session facilities (persistence, query, projections, storage, settings, credentials, telemetry), and the web host |
+| Host | one | The registries themselves (`tools`, `systemPrompt`, `agents`, `agent-loop`, `sessions`), cross-session facilities (persistence, query, projections, storage, settings, credentials, telemetry), the subagent providers those facilities resolve, and the web host |
 | Agent | one per session | What a single agent contributes to those registries: tool plugins, persona and prompt sections, compaction policy |

 Model routing stays out of presets. `installAgentLlmTarget` is already the per-agent seam for provider, model, and reasoning effort, and an LLM adapter mounted inside a preset would never be resolved by `agent-loop`, which lives in the host plane.
@@ -55,7 +55,7 @@ Which preset an unnamed session gets is a user setting (`agent-presets.default`)

 **Authoring a preset is an RPC, and a privileged one.** A composition is a file, but "edit it on the filesystem" is not a browser affordance, so the roster gained `read`/`write`/`remove` beside `select`. Those three are loopback-pinned: a composition names the plugins a session runs, so reading one is reconnaissance and writing one is arbitrary capability. `list` and `select` deliberately stay ordinary. The roster carries ids and trust only, and a LAN client's picker needs it; and choosing a preset looked like escalation — one of them mounts the toolset that edits the live runtime — but `session.create` already takes an `agentPreset`, so pinning only the switch would have left the same capability one method over. The capability is not the preset's to grant either: the deployment's own default already carries `bash` and the filesystem tools, so any caller that may start a session at all can already run commands as this process. Containment is a property of the id (`[a-z0-9][a-z0-9-]*`), checked before it becomes a directory name rather than by inspecting the joined path afterwards; the text is parsed with the loader's own schema and dialect, so a save cannot leave a file no session could load. Shipped presets are refused for writes and deletes, because the deployment's copy is what a broken local preset is compared against — which also makes "duplicate, then edit" the authoring path rather than an afterthought.

-**A service with a consumer outside the agent plane cannot move into a preset.** The aggressive split moved the `subagents` registry and its spawn/fork backends into the delegation group's entry-local realm, and `dsh web` then failed to boot: `dsh-host-apiproxy` is a HOST row that injects `subagents` to answer the browser's cross-session queries (`listChildren`, `followup`), so it waited forever for a service only sessions now provided. A per-session copy is wrong twice over — a provider name registers once, so the second session would have collided anyway. The registry and its backends are host-plane; the preset contributes the delegation TOOLS, which resolve the host registry. `workflows` stays entry-local because nothing outside an agent reads it. Grepping injectors is what should have caught this and did not: the search has to include the host packages, not just the agent-plane ones.
+**A service with a consumer outside the agent plane cannot move into a preset.** The aggressive split moved the `subagents` registry and its spawn/fork backends into the delegation group's entry-local realm, and `dsh web` then failed to boot: `dsh-host-apiproxy` is a HOST row that injects `subagents` to answer the browser's cross-session queries (`listChildren`, `followup`), so it waited forever for a service only sessions now provided. A per-session copy is wrong twice over — a provider name registers once, so the second session would have collided anyway. The registry and every shared backend, including the [fixed Codex and Claude Code product providers](2026-08-10-product-subagent-providers-in-shared-host.md), are host-plane; a preset contributes whichever delegation TOOLS its agent should see, and those tools resolve the host registry. `workflows` stays entry-local because nothing outside an agent reads it. Grepping injectors is what should have caught this and did not: the search has to include the host packages, not just the agent-plane ones.

 **A real-composition test that disables a host row cannot audit that row.** The web composition test disabled `api-gateway` — the api-proxy itself — as a row with side effects, which is exactly the row whose pending injection would have named the break. It now boots with the api-proxy enabled and the browse directory picker substituted, so the boot audit covers the whole host-plane injection graph; only the port, the asset tree, and the telemetry exporter stay off.

@@ -78,3 +78,7 @@ Which preset an unnamed session gets is a user setting (`agent-presets.default`)
 **Make the agent's scope key the preset.** Sessions on one preset would share a layer for free, but per-agent registrations — `installAgentLlmTarget`, per-agent tool restrictions — would then collide across sessions.

 **Run each preset as a child process.** [`subagent-dsh-sdk`](../../../../packages/subagent/subagent-dsh-sdk/README.md) already proves a full child harness works, and isolation would be absolute. It also means proxying streaming, approvals, and projections per session, which is a transport project rather than a composition one.
+
+**Give product subagents global enable settings and a separate settings page.** The process-wide value would compete with the preset as owner of model-visible tools and could not express two sessions using different compositions. Product providers stay host-side, while ordinary preset rows independently expose Codex and Claude Code tools.
+
+**Ship one preset for every Codex and Claude Code combination.** Four identities duplicate the full preset composition to represent two independent rows. A copied preset can enable either row directly, so combination presets add roster and maintenance cost without adding a user result.
@@ -18,7 +18,7 @@ Status: implemented

 | 平面 | 实例数 | 内容 |
 |---|---|---|
-| 宿主 | 一份 | 注册表本身（`tools`、`systemPrompt`、`agents`、`agent-loop`、`sessions`）、跨会话设施（持久化、查询、投影、存储、设置、凭据、遥测），以及 web 宿主 |
+| 宿主 | 一份 | 注册表本身（`tools`、`systemPrompt`、`agents`、`agent-loop`、`sessions`）、跨会话设施（持久化、查询、投影、存储、设置、凭据、遥测）、这些设施所解析的 subagent provider，以及 web 宿主 |
 | agent | 每会话一份 | 单个 agent 对这些注册表的贡献：工具插件、人设与提示词段落、压缩策略 |

 模型路由不进 preset。`installAgentLlmTarget` 已经是 provider、model 与 reasoning effort 的按 agent 可替换点；而挂在 preset 内部的 LLM 适配器永远不会被 `agent-loop` 解析到，因为后者位于宿主平面。
@@ -56,7 +56,7 @@ Status: implemented

 **创作 preset 是一次 RPC，而且是特权 RPC。** 组装是一个文件，但“去文件系统里改它”并不是浏览器能提供的操作，因此名单在 `select` 之外新增了 `read`/`write`/`remove`。这三者被固定在环回地址：组装指明了一个会话所运行的插件，因此读取它是侦察，写入它是任意能力。`list` 与 `select` 刻意保持为普通方法。名单只携带 id 与信任级别，而局域网客户端的选择器需要它；至于选择本身，它看起来像提权——其中一个 preset 会挂载可编辑活动运行时的工具集——但 `session.create` 本就接受 `agentPreset`，只固定切换会把同一能力留在隔壁一个方法上。这份能力也不由 preset 授予：部署自带的默认 preset 本就带着 `bash` 与文件系统工具，因此任何被允许开启会话的调用方，早已能以本进程的身份执行命令。约束是 id 自身的性质（`[a-z0-9][a-z0-9-]*`），在它成为目录名之前就检查，而不是事后再去审视拼接出的路径；文本使用 loader 自身的 schema 与方言解析，因此保存不会留下任何会话都无法加载的文件。随部署提供的 preset 拒绝写入与删除，因为部署自带的那一份正是用来对照有问题的本地 preset 的——这也让“先复制、再编辑”成为创作路径本身，而非事后补充。

-**在 agent 平面之外还有消费方的服务，不能搬进 preset。** 激进拆分把 `subagents` 注册表连同 spawn/fork 后端一起搬进了 delegation 组的 entry-local realm，于是 `dsh web` 直接起不来：`dsh-host-apiproxy` 是宿主行，它注入 `subagents` 来回答浏览器的跨会话查询（`listChildren`、`followup`），因而永远等待一个此刻只有会话才提供的服务。按会话各一份在两个层面上都是错的——provider 名只能注册一次，第二个会话本来也会相撞。注册表与后端属于宿主平面；preset 贡献的是委派**工具**，它们解析宿主注册表。`workflows` 保持 entry-local，因为 agent 之外没有任何东西读它。本该拦下它的是「检索注入方」这一步，而它没拦住：检索必须覆盖宿主包，而不只是 agent 平面的包。
+**在 agent 平面之外还有消费方的服务，不能搬进 preset。** 激进拆分把 `subagents` 注册表连同 spawn/fork 后端一起搬进了 delegation 组的 entry-local realm，于是 `dsh web` 直接起不来：`dsh-host-apiproxy` 是宿主行，它注入 `subagents` 来回答浏览器的跨会话查询（`listChildren`、`followup`），因而永远等待一个此刻只有会话才提供的服务。按会话各一份在两个层面上都是错的——provider 名只能注册一次，第二个会话本来也会相撞。注册表与所有共享后端，包括[固定的 Codex 与 Claude Code 产品 provider](2026-08-10-product-subagent-providers-in-shared-host.md)，都属于宿主平面；preset 只贡献自己的 agent 应看见的委派**工具**，这些工具解析宿主注册表。`workflows` 保持 entry-local，因为 agent 之外没有任何东西读它。本该拦下它的是「检索注入方」这一步，而它没拦住：检索必须覆盖宿主包，而不只是 agent 平面的包。

 **真实组装测试若禁用了某个宿主行，就无法审计该行。** web 组装测试把 `api-gateway`——也就是 api-proxy 本身——当作「有外部副作用的行」禁用了，而它恰恰是那个会以 pending 注入点名此次断裂的行。现在它在启用 api-proxy、并替换为 browse 目录选择器的前提下引导，启动审计因此覆盖整个宿主平面的注入图；只有端口、资源目录与遥测导出器仍然关闭。

@@ -79,3 +79,7 @@ Status: implemented
 **把 agent 的 scope 键设为 preset。** 同一 preset 上的会话就能免费共享一层，但按 agent 的注册——`installAgentLlmTarget`、按 agent 的工具限制——会跨会话相撞。

 **把每个 preset 作为子进程运行。** [`subagent-dsh-sdk`](../../../../packages/subagent/subagent-dsh-sdk/README.md) 已经证明完整的子 harness 可行，隔离性也会是绝对的。但这同时意味着要按会话代理流式输出、审批与投影，那是一个传输层项目，而非组装问题。
+
+**给产品 subagent 增加全局启用设置与独立设置页。** 进程级值会与 preset 争夺模型可见工具的所有权，也无法表达两个会话使用不同组装。产品 provider 留在宿主，普通 preset 行分别暴露 Codex 与 Claude Code 工具。
+
+**为 Codex 与 Claude Code 的每种组合交付一份 preset。** 四个身份会复制完整 preset 组装，只为表示两条独立行。复制后的 preset 已能直接启用任一行，因此组合 preset 只增加名单与维护成本，不增加用户结果。
@@ -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/2026-08-05-profile-plugin-bundles.md
-2026-08-05-profile-plugin-bundles.md: 2924b3cb445064fd47d82bcc94ec8d77ded5721b
-2026-08-05-profile-plugin-bundles.zh.md: b2287034010bcac1048bb385b2266f1bc75921da
+2026-08-05-profile-plugin-bundles.md: 54626e3f48a2ba7db19813e6e883f0e77499d0e2
+2026-08-05-profile-plugin-bundles.zh.md: 357e0f63d4eba0f0985c9e14aad54595c7b41c77
@@ -10,11 +10,9 @@ The `dsh` launcher hardcoded its compositions: `base.cordis.yml` + `web.cordis.y

 ## Decision

-Everything becomes a **profile**: a directory `$DSH_HOME/profiles/<name>` with a `package.json` (pnpm-managed out-of-tree plugin `dependencies` plus the profile manifest `dsh.profile` with its ordered `bundles` layer list) and a user `cordis.patch.yml`. A **bundle** is an npm package declaring `"dsh": { "bundle": { "patch": "./cordis.patch.yml" } }`; the two manifest kinds live under distinct `dsh.profile` / `dsh.bundle` keys so a package.json states which role it plays. The tree composes over an empty root by applying each bundle's patch in `dsh.profile.bundles` order, then the user layer, then `--patch` overlays, then flag patches — one `applyEntryPatches` call, identical for boot, flag derivation, and `--dump-config`.
+Everything becomes a **profile**: a directory `$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 and `--patch` overlays — one `applyEntryPatches` call shared by boot and `--dump-config`. App invocation values later moved from launcher-derived patches to startup services in the [app-owned command-line decision](2026-08-06-app-owned-command-line.md).

-The shipped bundles are `@deepseek-ai/dsh-base` (shared core rows), `@deepseek-ai/dsh-web-app` (browser Host rows and Web runtime glue), and `@deepseek-ai/dsh-headless` (a direct one-shot runner over base, without web-app). `dsh web` is the Web-flag alias for `--profile web`; `dsh run [--profile <name>] "task"` owns one-shot execution and defaults to the headless profile; generic `dsh --profile <name>` boots without a task. Patch overlays use `--patch`. `dsh plugin --profile <name> <args...>` is a thin pnpm forwarder that initializes the profile and reconciles `dsh.profile.bundles` with installed bundle declarations; a package without a bundle declaration remains a plain dependency. [Headless as a direct core entry point](2026-08-09-headless-direct-core-entry-point.md) owns the headless composition contract.
-
-The [`dsh run` command decision](../feature/2026-08-08-dsh-run-headless-command.md) owns the one-shot grammar; this note owns the profile composition it selects.
+The shipped bundles are `@deepseek-ai/dsh-base` (shared core rows), `@deepseek-ai/dsh-web-app` (browser Host rows and Web runtime glue), and `@deepseek-ai/dsh-headless` (a direct one-shot runner over base, without web-app). Generic `dsh --profile <name>` hands its remaining arguments to that profile's command-line startup row: Web owns its flag family, while headless owns its task positional. Patch overlays use launcher-owned `--patch`. `dsh plugin --profile <name> <args...>` is a thin pnpm forwarder that initializes the profile and reconciles `dsh.profile.bundles` with installed bundle declarations; a package without a bundle declaration remains a plain dependency. [Headless as a direct core entry point](2026-08-09-headless-direct-core-entry-point.md) owns the headless composition contract.

 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).

@@ -24,7 +22,7 @@ Two supporting refactors: the webserver's built-in static dist serving became th

 - **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` host hook is the one host-provided slot, and it is provided in `boot()`'s `prepare` hook, before any config-tree entry mounts.
+- **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 and app-owned startup services, so the composition stays fully dumpable and the manifest stays data-only. The launcher-owned `ctx.headlessIo` host hook 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
@@ -10,11 +10,9 @@ Status: implemented

 ## Decision

-一切都变成 **profile**：即目录 `$DSH_HOME/profiles/<name>`，其中包含一个 `package.json`（pnpm 管理的树外插件 `dependencies`，加上 profile manifest（元数据清单）`dsh.profile` 及其有序的 `bundles` 层列表）和一份用户 `cordis.patch.yml`。**组合包**（bundle）是声明了 `"dsh": { "bundle": { "patch": "./cordis.patch.yml" } }` 的 npm 包；两种 manifest 分别位于互不相同的 `dsh.profile` / `dsh.bundle` 键下，因此一份 package.json 能说明自己扮演哪种角色。配置树在空的根之上组合：按 `dsh.profile.bundles` 顺序应用每个组合包的 patch，然后是用户层，然后是 `--patch` overlay，最后是 flag patch——全部收敛为一次 `applyEntryPatches` 调用，启动、flag 派生与 `--dump-config` 使用完全相同的路径。
+一切都变成 **profile**：即目录 `$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，然后是用户层与 `--patch` overlay——启动与 `--dump-config` 共享同一条 `applyEntryPatches` 路径。随后，[应用持有命令行的决策](2026-08-06-app-owned-command-line.md)又把调用期取值从启动器派生的 patch 迁移到了启动服务。

-随附的组合包是 `@deepseek-ai/dsh-base`（共享核心配置行）、`@deepseek-ai/dsh-web-app`（浏览器 Host 配置行与 Web 运行时粘合层）和 `@deepseek-ai/dsh-headless`（直接叠加在 base 上且不含 web-app 的一次性 runner）。`dsh web` 是携带 Web flag 家族的 `--profile web` 别名；`dsh run [--profile <name>] "task"` 负责一次性执行，默认使用 headless profile；通用的 `dsh --profile <name>` 启动 profile 而不携带任务。patch overlay 使用 `--patch`。`dsh plugin --profile <name> <args...>` 是一层薄薄的 pnpm 转发器，负责初始化 profile，并依据已安装包的组合包声明调和 `dsh.profile.bundles`；没有组合包声明的包保持为普通依赖。[Headless 作为直接 core 入口](2026-08-09-headless-direct-core-entry-point.md)负责 headless 组合约定。
-
-[`dsh run` 命令决策](../feature/2026-08-08-dsh-run-headless-command.md)负责一次性语法；本 Agent Note 负责该语法所选择的 profile 组合。
+随附的组合包是 `@deepseek-ai/dsh-base`（共享核心配置行）、`@deepseek-ai/dsh-web-app`（浏览器 Host 配置行与 Web 运行时粘合层）和 `@deepseek-ai/dsh-headless`（直接叠加在 base 上且不含 web-app 的一次性 runner）。通用的 `dsh --profile <name>` 把剩余参数交给该 profile 的命令行启动行：Web 持有自己的 flag 家族，headless 则持有任务位置参数。patch overlay 使用启动器持有的 `--patch`。`dsh plugin --profile <name> <args...>` 是一层薄薄的 pnpm 转发器，负责初始化 profile，并依据已安装包的组合包声明调和 `dsh.profile.bundles`；没有组合包声明的包保持为普通依赖。[Headless 作为直接 core 入口](2026-08-09-headless-direct-core-entry-point.md)负责 headless 组合约定。

 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）。

@@ -24,7 +22,7 @@ Status: implemented

 - **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 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` 宿主钩子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。
+- **在组合包 manifest 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——粘合逻辑就是普通配置行和由应用持有的启动服务，因此组合始终可完整 dump，manifest 保持纯数据。启动器持有的 `ctx.headlessIo` 宿主钩子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
@@ -1,6 +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/2026-08-08-dsh-run-headless-command.md
-2026-08-08-dsh-run-headless-command.md: ed095f4077a23e51bffb647d24eed19ba09e11ed
-2026-08-08-dsh-run-headless-command.zh.md: 89d54e35573f14786e05d648f2b42891ca27a043
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-08-06-app-owned-command-line.md
+2026-08-06-app-owned-command-line.md: 4a05cac5ed7f44fb55c2d4498bf28a43befdb073
+2026-08-06-app-owned-command-line.zh.md: 86a37f416d17c4615152b29d73f171803f24c4c3
@@ -0,0 +1,51 @@
+# Agent Note: Apps own their command line through `ctx.cmdlineArgs`
+
+Status: implemented
+
+English | [中文](2026-08-06-app-owned-command-line.zh.md)
+
+## Problem
+
+After profiles, compositions were installable but their command lines were not. `apps/cli` still declared the Web flag family (`--host`, `--port`, `--dev`, `--workspace-root`, `--trusted-host`) and the one-shot task positional, then derived patches for row ids it hardcoded (`webserver`, `api-gateway`, `connection`, `web-runtime`). An out-of-tree app such as [turtle-ui](https://github.com/deepseek-harness/turtle-ui) could contribute rows but had no way to accept a flag: `dsh --profile tui --resume <session>` had nowhere to be parsed, and `dsh --profile web --help` printed the launcher's help rather than the web app's.
+
+## Decision
+
+The launcher parses only what it owns — `--profile`, `--patch`, the config dumps — and hands **everything after its own flags** to the booted tree verbatim. The split is positional: the first token the launcher does not recognize starts the app's arguments (commander's `passThroughOptions` + `allowUnknownOption` + `helpOption(false)`). A bare `dsh -h`, which has no app to hand the flag to, still prints the launcher's own help.
+
+The new `@deepseek-ai/dsh-cmdline` package owns the handoff. A launcher calls `provideCmdline(ctx, host)` before any entry mounts, providing `ctx.cmdlineArgs` (whose whole interface is `get(): readonly string[]`) and `ctx.appExit`. Any ordinary app plugin may inject `cmdlineArgs`, call `parseCmdline(ctx, program, plan)` with its own commander program, and provide the returned value as an app-owned service. Its Loader row carries no launcher marker or special kind, and the launcher does not inspect the composition for an owner. Multiple plugins may read the same immutable snapshot; a profile with no reader ignores its app arguments. Rows configured from a provider inject its service and read direct lazy config expressions (`port: !!js ctx.webStartup.port ?? 3080`), so a flag beats the value written beside it and nothing is written back into any row.
+
+The boot mounts the composition once. Cordis holds each row until its injections are active; Loader then interpolates that row's `!!js` against the injection-ready plugin context immediately before activation. Include keeps nested row expressions raw until their target row reaches this point. `--help` leaves the provider's service absent, so dependent rows never activate, and a live patch reload interpolates again against the service that remains active, so a served port cannot be silently reset.
+
+The shipped apps moved their flags into their bundles: `dsh-web-app` owns the Web family (and enables the `client-hmr` row it now ships disabled, for `--dev`), and `dsh-headless` owns the task positional and rejects a missing task as a usage error. `apps/cli/src/web.ts` is gone; `runProfile` no longer knows any flag-target row id. Out of tree, turtle-ui gained `--resume <session>` / `--session <id>` the same way, which is the design's real validation: an installed plugin added a flag with no launcher change.
+
+Two further consequences. Loader mounts sibling rows concurrently, so one row can activate while another still mounts or while the whole boot is rolling back; the Web bundle therefore publishes its URL only after its own Loader tree settles. The Web bundle's runtime plugin owns the harness-source prompt section too, so `dsh web` and `dsh --profile web` boot identically without Web-specific launcher setup.
+
+## Why Loader owns the ordering
+
+Four framework facts shape the mechanism:
+
+- **A profile's rows arrive inside the root include's `patches` option.** Include is an entry-tree owner, so its static entry-config resolver interpolates Include's own options while preserving nested `!!js` nodes for their target rows instead of recursively evaluating them in the Include context.
+- **Cordis activates a fiber only after all declared injections are active.** Immediately before each activation, Cordis runs the `internal/config` waterfall against the fiber's own context; Loader's listener interpolates the raw config after Cordis snapshots its injected services.
+- **Provider replacement and HMR must preserve the same contract.** Fiber reactivation re-runs the waterfall, HMR carries the raw config to the replacement fiber, and a pending row accepts option changes without prematurely evaluating expressions against absent services.
+- **A row cannot be inserted from inside a mounting plugin** — `tree.create` returns a prefixed id it then fails to resolve — so a conditional row ships `disabled: true` and an active row enables it (`dsh web --dev` and its reload chain). Enablement is an in-memory Loader override rather than an options rewrite, so Include reapplication cannot silently disable it. The Web bundle also starts client discovery only after enabling the optional row, ensuring the first browser graph already contains its HMR receiver.
+
+This leaves dependency ordering in Cordis activation and Loader interpolation, which own it. Rows keep their `inject` and config, Loader mounts the composition once, and the launcher only provides argv and process-lifecycle services.
+
+## Alternatives considered
+
+- **Writing the resolved values into each row** (a config update per row, plus a patch layer handed back to the launcher so a reload could not undo it): it worked, but it meant patches travelling from an app to the launcher and back, two mechanisms for one fact, and a recycle whose correctness depended on Loader restart internals. The maintainer rejected the round trip; the service the rows read replaced all of it.
+- **Releasing rows by clearing their `inject`**: it worked in isolation and failed on the real web tree, because clearing `inject` is exactly what loses the plugin's static injections. The failure is silent until a plugin reads a service it declared.
+- **Launcher-managed two-pass mounting**: it can make a provider active before readers are applied, but duplicates the composition, makes ordering a launcher concern, and conceals the Loader defect that nested expressions were evaluated in the include context rather than the target row's injected context.
+- **The launcher running each bundle's command function before boot** (no Cordis involvement): strictly earlier than "boot, then help", but it makes app startup a second plugin protocol outside the tree. An ordinary `cmdlineArgs`-injected provider keeps one protocol and remains dumpable and patchable.
+- **A launcher-enforced command-line owner**: rejecting zero or multiple readers would arbitrate overlaps such as `-h`, but `get()` is an immutable read and normal composition may need several app-owned services. Plugins therefore share the snapshot and own any parser interaction through ordinary composition.
+- **`instanceof CommanderError`**: an out-of-tree plugin brings its own commander copy, so the class identity differs and a printed `--help` was rethrown as a fatal load failure. Commander's control-flow errors are detected structurally instead.
+
+## Consequences
+
+- An app's flags, help text, and usage errors live with the rows they configure; adding a flag to an installed plugin needs no launcher change.
+- The launcher still recognizes the headless runner for one-shot process lifetime and the telemetry row for its environment switch; neither path interprets app arguments.
+- `--help` leaves every row that depends on the provider's service pending and requests bounded exit; unrelated rows may activate concurrently before teardown.
+- An app-owned service has no statically declared provider: a bundle shipping consumer rows without that provider fails at settlement with pending entries naming the service, not at load.
+- A user patch that replaces a row's whole `config` drops its expressions, and with them the flag's precedence for that row.
+- Launcher flags must precede app arguments; a first app argument equal to `web` or `plugin` selects that subcommand instead, `-V`/`--version` remains launcher-owned before that boundary, and the launcher's parser consumes one `--`, so a literal `--` for the app needs `-- --`.
+- `--dump-config` never runs app command-line providers, so it prints the composition before any app argument is resolved and rejects an invocation that carries app arguments.
@@ -0,0 +1,51 @@
+# Agent Note: 应用通过 `ctx.cmdlineArgs` 持有自己的命令行
+
+Status: implemented
+
+[English](2026-08-06-app-owned-command-line.md) | 中文
+
+## 问题
+
+profile 落地之后，组合可以安装，命令行却不能。`apps/cli` 仍然声明着 Web flag 家族（`--host`、`--port`、`--dev`、`--workspace-root`、`--trusted-host`）和一次性任务位置参数，再为自己硬编码的行 id（`webserver`、`api-gateway`、`connection`、`web-runtime`）派生 patch。像 [turtle-ui](https://github.com/deepseek-harness/turtle-ui) 这样的树外应用能贡献行，却无处接受一个 flag：`dsh --profile tui --resume <session>` 没有地方可供解析，而 `dsh --profile web --help` 打印的是启动器的 help，而不是 web 应用的 help。
+
+## 决策
+
+启动器只解析属于自己的部分（`--profile`、`--patch`、配置 dump），并把**自己 flag 之后的一切**原样交给引导起来的配置树。切分按位置进行：启动器不认识的第一个 token 就是应用参数的起点（依靠 commander 的 `passThroughOptions` + `allowUnknownOption` + `helpOption(false)`）。裸的 `dsh -h` 没有可交付的应用，仍然打印启动器自己的 help。
+
+新包 `@deepseek-ai/dsh-cmdline` 持有这次交接。启动器在任何条目挂载之前调用 `provideCmdline(ctx, host)`，提供 `ctx.cmdlineArgs`（其全部接口就是 `get(): readonly string[]`）与 `ctx.appExit`。任何普通应用插件都可以注入 `cmdlineArgs`，用自己的 commander program 调用 `parseCmdline(ctx, program, plan)`，再把返回值作为应用自有服务提供出去。它的 Loader 行不携带启动器标记或特殊类型，启动器也不会检查组合中的所有者。多个插件可以读取同一份不可变快照；没有读取方的 profile 会忽略自己的应用参数。由提供方配置的行注入其服务，并在惰性配置表达式中直接读取它（`port: !!js ctx.webStartup.port ?? 3080`），因此 flag 胜过写在它旁边的值，也没有任何东西被写回任何一行。
+
+boot 只挂载一次整套组合。Cordis 让每一行等待其注入激活；Loader 随后在激活前一刻，基于已注入就绪的插件上下文插值该行的 `!!js`。Include 会保留嵌套的行表达式，直到目标行到达这一时点。`--help` 会让提供方服务保持缺失，因此依赖行永不激活；活动 patch 重载会针对仍然在线的服务再次插值，所以已经服务中的端口不会被悄悄重置。
+
+已交付的各应用把自己的 flag 搬进了组合包：`dsh-web-app` 持有 Web 家族（并为 `--dev` 启用它如今以禁用状态交付的 `client-hmr` 行），`dsh-headless` 持有任务位置参数，缺少任务时按用法错误拒绝。`apps/cli/src/web.ts` 已删除；`runProfile` 不再知道任何 flag 目标行 id。在树外，turtle-ui 以同样的方式获得了 `--resume <session>` / `--session <id>`，这才是这套设计的真正验证：一个已安装的插件加上了一个 flag，启动器毫无改动。
+
+还有两条后果。Loader 会并发挂载兄弟行，因此一行可能已经激活，而另一行仍在挂载，或整次 boot 正在回滚；所以 Web 组合包只会在自身的 Loader 配置树结算后公布 URL。另外，Web 组合包的运行时插件也持有 harness 源码提示词段，因此 `dsh web` 与 `dsh --profile web` 无需 Web 专用启动器设置即可按完全相同的方式启动。
+
+## 为什么由 Loader 持有顺序
+
+四条框架事实塑造了这套机制：
+
+- **profile 的各行位于根 include 的 `patches` 选项内部。** Include 是条目树所有者，因此它的静态条目配置解析器会插值 Include 自身的选项，同时为目标行保留嵌套的 `!!js` 节点，而不是在 Include 上下文中递归求值。
+- **Cordis 只在所有声明的注入都已激活后才激活 fiber。** 每次激活前一刻，Cordis 会基于 fiber 自身上下文运行 `internal/config` waterfall；Cordis 快照注入服务之后，Loader 的监听器再插值原始配置。
+- **提供方替换与 HMR 必须保持相同契约。** fiber 重新激活时会重跑 waterfall，HMR 会把原始配置带给替换 fiber，而待处理行可以接受选项变更，不会针对缺失服务提前求值表达式。
+- **不能从正在挂载的插件内部插入一行**——`tree.create` 返回一个带前缀的 id，随后它自己解析不出来——因此条件性的行以 `disabled: true` 交付，再由活跃行启用（`dsh web --dev` 及其重载链路）。启用采用 Loader 的内存覆盖而非改写选项，因此 Include 重新应用配置时不会悄然将其禁用。Web 组合包还会在启用可选行之后才启动客户端发现，确保首份浏览器图中已经包含 HMR 接收端。
+
+这样，依赖顺序仍由负责它的 Cordis 激活与 Loader 插值流程处理。各行保留自己的 `inject` 和配置，Loader 只挂载一次组合，启动器只提供 argv 与进程生命周期服务。
+
+## 曾考虑的替代方案
+
+- **把解析出的取值写进每一行**（逐行一次配置更新，外加交还给启动器的一层 patch，使重载无法撤销它）：它能工作，但这意味着 patch 在应用与启动器之间来回传递、同一件事有两套机制，以及一套其正确性依赖 Loader 重启内部细节的回收重建。维护者否决了这次往返；供各行读取的服务取代了这一切。
+- **通过清空行的 `inject` 来放行**：孤立测试可行，在真实 web 树上失败，因为清空 `inject` 恰恰会丢失插件的静态注入。在插件真的去读它声明过的服务之前，这个失败是静默的。
+- **由启动器管理两趟挂载**：它可以让提供方先于读取行激活，但会重复组合、把顺序变成启动器职责，还掩盖了 Loader 的缺陷——嵌套表达式在 include 上下文而不是目标行的注入上下文中求值。
+- **由启动器在 boot 之前运行每个组合包的命令函数**（完全不经过 Cordis）：严格早于「先 boot 再 help」，但这会让应用启动成为配置树之外的第二套插件协议。使用注入 `cmdlineArgs` 的普通提供方只保留一套协议，并且仍可 dump、可 patch。
+- **由启动器强制指定命令行所有者**：拒绝零个或多个读取方可以裁决 `-h` 等重叠项，但 `get()` 是不可变读取，普通组合也可能需要多个应用自有服务。因此插件共享该快照，并通过普通组合持有各自解析器的交互。
+- **`instanceof CommanderError`**：树外插件会带来自己的一份 commander 副本，类身份因此不同，已经打印出来的 `--help` 会被重新抛成致命的加载失败。改为按结构识别 commander 的控制流错误。
+
+## 后果
+
+- 应用的 flag、help 文本和用法错误与它们所配置的行放在一起；给已安装的插件加一个 flag 不需要改动启动器。
+- 启动器仍会识别 headless runner 以管理一次性进程生命周期，并识别 telemetry 行以应用环境开关；两条路径都不解析应用参数。
+- `--help` 会让所有依赖提供方服务的行保持待处理并请求有边界的退出；无关行可能在拆除前并发激活。
+- 应用自有服务没有静态声明的提供方：交付了消费行却缺少对应提供方的组合包会在结算时失败，报出指向该服务的待处理条目，而不是在加载时失败。
+- 用户 patch 若整体替换某行的 `config`，会连同其中的表达式一起丢掉，该行上 flag 的优先级也随之消失。
+- 启动器的 flag 必须写在应用参数之前；如果应用的第一个参数恰好等于 `web` 或 `plugin`，会选择对应的子命令；`-V`／`--version` 在该边界之前仍归启动器持有；而且启动器的解析器会消耗掉一个 `--`，因此要给应用传一个字面量 `--` 需要写成 `-- --`。
+- `--dump-config` 从不运行应用命令行提供方，因此它在任何应用参数被解析之前打印组合，并拒绝携带应用参数的调用。
@@ -1,51 +0,0 @@
-# Agent Note: Trusted repository packages load Cordis code
-
-Status: implemented
-
-English | [中文](2026-08-08-trusted-repository-package-code.zh.md)
-
-## Problem
-
-The standalone repository format already installs a selected Git package and runs its dependency and lifecycle code with host authority, but it exposed only copied skills and MCP metadata to DSH. Forbidding a Cordis entry did not create a security boundary: package installation remained trusted executable code while the restriction prevented the package from contributing the Plugin behavior that the Harness architecture is designed to compose.
-
-A repository author also needs to keep an ordinary TypeScript npm package shape. Requiring publication to npm, pre-generated JavaScript in Git, or a DSH-owned TypeScript compiler would make a Git source less capable than the same package installed through a developer-owned SDK project. The first model request must observe any MCP tools that this package starts; background-only initial discovery makes a successful installation nondeterministic at the application boundary.
-
-## Decision
-
-A configured repository package is trusted code. Its `.dsh-plugin/package.json` may declare `dsh.entry` as a relative path to a compiled ESM Cordis Plugin inside that package, alongside or instead of `dsh.skills` and `dsh.mcpServers`. At least one contribution is required. The entry may use namespace exports or a default export and retains ordinary Cordis semantics for `name`, `inject`, `Config`, registrations, startup failure, and effect-scoped teardown.
-
-The package owns its npm dependencies and build toolchain. It declares the published `@deepseek-ai/dsh-repository-plugin` package to obtain the `dsh-plugin-prepare` executable. `scripts.prepack` is a non-empty package-authored command that must invoke that dependency-provided helper, but it may first run `tsc`, `tsdown`, or any other build. DSH neither injects the helper, parses the shell program, nor compiles repository source. The helper validates the metadata after the preceding build, requires the configured entry to resolve to a file within `.dsh-plugin`, validates and copies declared static assets, and writes the prepared `dsh-plugin.mjs` wrapper. The installed package must retain a `prepack` declaration containing that helper command; a missing dependency, wrapper, or build output fails before a cache generation becomes usable.
-
-The generated wrapper first mounts the DSH-owned static runtime for skills and MCP definitions, then dynamically imports and unwraps the explicit entry and mounts it as a child. The wrapper statically declares dependencies implied by the prepared manifest; an entry module's additional `inject` is discovered only when mounted and must already be available in the host composition. Both children must reach Cordis `ACTIVE`; an unsatisfied `inject` or startup exception rejects the repository Loader transaction instead of committing an inert generation. Loader removal, failed replacement, and parent disposal unwind the entry, skill providers, MCP clients, and their effects together.
-
-`dsh-mcp-client` resolves its initial connection and tool synchronization promise as part of Plugin application. Its entry is an `async function`, not an ordinary function returning a Promise: Cordis identifies prototype-bearing ordinary functions as constructors and does not treat a constructor's returned Promise as startup work. A valid server's tools therefore exist before its parent repository wrapper activates and before a one-shot application starts its first model request. Its `failOnStartupError` config preserves optional standalone servers by default while letting repository adapters require their declared servers. Repository-translated MCP clients enable that mode, so initial connection, discovery, or tool-registration failure rejects the candidate generation and rollback still closes the transport.
-
-## Trust boundary
-
-Exact refs, source containment, credential-shaped environment scrubbing, prepared manifests, and immutable cache keys protect identity and composition integrity; they do not sandbox executable package input. Repository lifecycle scripts, transitive npm dependencies, the compiled entry, and spawned MCP servers can exercise the authority available to the DSH process and the Cordis services they receive. Users must therefore trust the selected repository and should pin immutable refs and grant Git only the narrow read credential needed for acquisition.
-
-Model-visible behavior remains governed by the owning DSH seam. A repository entry may register tools, prompt sections, policies, commands, agents, or other effects, but anything reaching a model request still needs the corresponding logged DSH representation and lifecycle cleanup. The repository format grants code loading; it does not weaken those service contracts.
-
-## Alternatives considered
-
-**Keep code forbidden while allowing arbitrary package lifecycles.** Rejected because installation already executes trusted repository code, so the restriction added no isolation and forced Plugin authors to publish or maintain a second integration path.
-
-**Have DSH compile repository TypeScript.** Rejected because compiler choice, module layout, generated chunks, native dependencies, and package metadata belong to the npm package. Running the package's declared build preserves the same boundary as other Git dependencies.
-
-**Import `main`, `exports`, or another discovered entry implicitly.** Rejected because an npm package may contain utilities or an MCP executable that is not a Cordis Plugin. The explicit `dsh.entry` field makes code activation reviewable and lets preparation validate the packed path.
-
-**Add a closed manifest field for every future DSH contribution.** Rejected as the universal extension mechanism. Skills and common MCP files retain useful portable static adapters, while DSH-native behavior composes through the existing Cordis Plugin and service contracts.
-
-## Consequences
-
- A TypeScript DSH Plugin can live in a GitHub repository, install ordinary npm dependencies, compile during `prepack`, and run without publishing the Plugin package to npm.
- Static-only repository packages remain valid and retain import-free wrappers; adding `dsh.entry` opts that package into runtime code import.
- A package build, dependency install, entry import, unmet service, or Plugin startup failure prevents the candidate generation from replacing the last good configuration.
- Initial MCP synchronization can lengthen application startup by the MCP SDK's per-request timeout, and a repository-declared server that is unavailable or cannot publish its complete tool generation prevents that candidate generation from activating.
- Repository code receives host authority, so source review and immutable pinning are operational security requirements rather than optional hardening.
-
-## Testing
-
-Repository-format tests prepare and mount default-export code entries through the real Loader, observe an entry-owned service, remove the Loader row, and observe cleanup; they also retain skill/MCP preparation, containment, damaged-package, pending-service, and rollback coverage. MCP lifecycle tests require `apply` to settle only after initial tool publication, preserve opt-in contained startup failure, and prove strict connection or tool-registration rejection still closes the client.
-
-The Node 24 consumer acceptance uses the actual built `dsh run` command with a fresh DSH home and an authenticated private GitHub source pinned to the pull request's exact head SHA. The test packs the current repository Plugin build with the same private-field removal and workspace-dependency pinning used for publication, serves its packument and tarball from a job-local npm registry, and directs the Git package's ordinary scoped npm resolution there. That repository package obtains `dsh-plugin-prepare` from the simulated published dependency, installs its other pinned runtime and development dependencies, type-checks and bundles TypeScript during `prepack`, prepares a skill plus a stdio MCP server and `dsh.entry`, exposes the skill and MCP schema in the first real model request, executes the MCP tool, and lets the compiled Cordis entry append a second marker to the result observed in the following request. Registry and cache assertions require npm resolution to reach the simulated publication, source files to be absent from the packed installation, and both built modules, their installed dependency, copied assets, and generated wrapper to be present.
@@ -1,51 +0,0 @@
-# Agent Note: 受信任 repository 包加载 Cordis 代码
-
-状态：已实现
-
-[English](2026-08-08-trusted-repository-package-code.md) | 中文
-
-## 问题
-
-独立 repository 格式已经会安装选定的 Git 包，并以宿主权限运行其依赖和生命周期代码，但它向 DSH 暴露的只有复制后的 skill（技能）和 MCP 元数据。禁止 Cordis 入口并未建立安全边界：包安装过程仍会执行受信任代码，而这项限制却阻止包贡献 Harness 架构本就用于组合的插件行为。
-
-仓库作者还需要保持普通 TypeScript NPM 包的结构。如果要求发布到 NPM、把预生成的 JavaScript 签入 Git，或使用 DSH 自有的 TypeScript 编译器，Git 源的能力就会弱于通过开发者自有 SDK 项目安装的同一个包。首个模型请求必须看到该包启动的所有 MCP 工具；仅在后台进行初始发现，会让一次成功安装在应用边界上具有不确定性。
-
-## 决策
-
-已配置的 repository 包是受信任代码。其 `.dsh-plugin/package.json` 可以连同 `dsh.skills` 和 `dsh.mcpServers` 声明 `dsh.entry`，也可以用它取代二者；`dsh.entry` 是指向该包内已编译 ESM Cordis 插件的相对路径。至少需要一种贡献。入口可以使用 namespace 导出或 default export，并沿用 Cordis 对 `name`、`inject`、`Config`、注册、启动失败和 effect 作用域清理的常规语义。
-
-包自行负责其 NPM 依赖和构建工具链。它声明已发布的 `@deepseek-ai/dsh-repository-plugin` 包以取得 `dsh-plugin-prepare` 可执行文件。`scripts.prepack` 是由包作者编写的非空命令，必须调用该依赖提供的辅助程序，但可以先运行 `tsc`、`tsdown` 或其他任意构建。DSH 不会注入辅助程序，也不会解析该 shell 程序或编译 repository 源码。辅助程序会在前序构建之后校验元数据，要求已配置入口解析到 `.dsh-plugin` 内的文件，校验并复制已声明的静态资源，再写入已准备的 `dsh-plugin.mjs` 包装层。已安装包必须保留包含该辅助命令的 `prepack` 声明；依赖、包装层或构建输出缺失会在缓存 generation 可用前导致失败。
-
-生成的包装层先挂载 DSH 自有的静态运行时来处理 skill 和 MCP 定义，再动态导入显式入口、解包其导出并将其挂载为子级。包装层会静态声明已准备 manifest（元数据清单）所隐含的依赖；入口模块的额外 `inject` 只有在挂载时才会被发现，并且此时必须已存在于宿主组合中。两个子级都必须进入 Cordis `ACTIVE`；无法满足的 `inject` 或启动异常会拒绝 repository Loader 事务，而不会提交未激活的 generation。Loader 移除、替换失败和父级 dispose（资源释放）会一并撤销入口、skill 提供方、MCP client 及其 effect。
-
-`dsh-mcp-client` 会在插件应用期间完成其初始连接和工具同步 promise。其入口必须是 `async function`，而不是返回 Promise 的普通函数：Cordis 会把带 prototype 的普通函数识别为 constructor，不会把 constructor 返回的 Promise 当作启动工作。因此，有效 server 的工具会在父级 repository 包装层激活前、一次性应用发起首个模型请求前就已存在。其 `failOnStartupError` 配置默认保留独立可选 server 的行为，同时允许 repository adapter 要求已声明 server 必须可用。Repository 转换出的 MCP client 会启用该模式，因此初始连接、发现或工具注册失败会拒绝候选 generation，回滚仍会关闭 transport。
-
-## 信任边界
-
-精确 ref、源路径包含约束、清除名称符合凭据模式的环境变量、已准备的 manifest 和不可变缓存键，可以保护身份与组合完整性；它们不会为可执行包输入提供沙箱隔离。Repository 生命周期脚本、传递性 NPM 依赖、已编译入口和 spawn 的 MCP server 可以行使 DSH 进程可用的权限，以及它们所获 Cordis 服务授予的权限。因此，用户必须信任所选仓库，应当固定不可变 ref，并只授予 Git 获取源码所需的最小只读凭据。
-
-模型可见行为仍由所属 DSH seam 管理。repository 入口可以注册工具、提示词段落、策略、命令、agent（智能体）或其他 effect，但任何进入模型请求的内容仍须具有对应的 DSH 日志表示和生命周期清理。repository 格式授予代码加载能力；它不会削弱这些服务约定。
-
-## 考虑过的替代方案
-
-**继续禁止代码，但允许任意包生命周期。** 拒绝，因为安装过程本就执行受信任的 repository 代码，所以该限制没有提供隔离，反而迫使插件作者发布或维护第二条集成路径。
-
-**由 DSH 编译 repository TypeScript。** 拒绝，因为编译器选择、模块布局、生成分片、原生依赖和包元数据属于 NPM 包。运行包所声明的构建，可以保持与其他 Git 依赖相同的边界。
-
-**隐式导入 `main`、`exports` 或其他发现的入口。** 拒绝，因为 NPM 包可能包含并非 Cordis 插件的实用工具或 MCP 可执行文件。显式 `dsh.entry` 字段使代码激活可供评审，并让准备阶段校验打包后的路径。
-
-**为未来每种 DSH 贡献添加封闭 manifest 字段。** 不采用它作为通用扩展机制。skill 和通用 MCP 文件仍保留有用的可移植静态适配器；DSH 原生行为则通过现有 Cordis 插件与服务约定组合。
-
-## 后果
-
- TypeScript DSH 插件可以存放在 GitHub 仓库中，安装普通 NPM 依赖，在 `prepack` 期间完成编译，并在无需把插件包发布到 NPM 的情况下运行。
- 仅含静态贡献的 repository 包仍然有效，并保留无 import 包装层；添加 `dsh.entry` 会使该包选择启用运行时代码导入。
- 包构建、依赖安装、入口导入、所需服务未满足或插件启动失败，都会阻止候选 generation 替换最后一个可用配置。
- 初始 MCP 同步可能因 MCP SDK 的单次请求超时而延长应用启动时间；repository 声明的 server 不可用或无法发布完整工具 generation 时，该候选 generation 无法激活。
- Repository 代码获得宿主权限，因此源码评审和锁定不可变 ref 是运行安全要求，而不是可选加固措施。
-
-## 测试
-
-repository 格式测试通过真实 Loader 准备并挂载使用 default export 的代码入口，观察入口自有服务，移除 Loader 配置项，再观察清理；测试还保留针对 skill／MCP 准备、路径包含约束、包损坏、等待服务和回滚的覆盖。MCP 生命周期测试要求 `apply` 只在初始工具发布后完成，保留可选择启用的启动失败收束行为，并证明严格连接拒绝或工具注册拒绝仍会关闭 client。
-
-Node 24 消费方验收使用实际构建的 `dsh run` 命令、全新 DSH 主目录，以及锁定到 PR（Pull Request）的精确 head SHA 且经过认证的私有 GitHub 源。测试会采用发布时相同的移除 `private` 字段和固定 workspace 依赖版本流程，对当前 repository 插件构建进行打包；再由作业本地 NPM 注册表提供其 `packument` 与 tarball，并把 Git 包的常规 scoped NPM 解析指向该注册表。该 repository 包从模拟发布的依赖取得 `dsh-plugin-prepare`，安装其他固定版本的运行时依赖与开发依赖，在 `prepack` 期间对 TypeScript 进行类型检查和打包，准备一个 skill、一个 stdio MCP server 及 `dsh.entry`，在首个真实模型请求中暴露 skill 与 MCP schema，执行 MCP 工具，并让已编译 Cordis 入口向结果追加第二个标记，供后续请求观察。注册表与缓存断言要求 NPM 解析必须命中模拟发布，打包安装中不存在源码文件，同时必须存在两个已构建模块、其已安装依赖、复制资源和生成包装层。
@@ -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/2026-08-09-client-conversation-node-assembly.md
-2026-08-09-client-conversation-node-assembly.md: 16a39539064644e5467f701789a7e2ef1f7ff172
-2026-08-09-client-conversation-node-assembly.zh.md: 0e0fbdf8f3320393022528e6e3fe2cf0d492a1d3
+2026-08-09-client-conversation-node-assembly.md: f6cd7ea94d485d92fd4ea178751c08bd01ecb4b2
+2026-08-09-client-conversation-node-assembly.zh.md: e550b7870a611ec625d7c2a738bf71947825ea58
@@ -148,7 +148,7 @@ The Assembler verifies `node.key === context.key` and `node.target === target`.

 `current` lets a Definition distinguish "never materialized" from "already materialized and now hidden." Assistant retry and Turn Error suppression use it to avoid illegal Node withdrawal.

-A Definition may branch by target to construct different data, while matching, Context identity, and State remain target-neutral. This change registers only the `chat` builder; Trajectory remains on its independent `session-history` fold until it gains a registered target.
+A Definition owns at most one view target; state-only Definitions omit both `target` and `buildViewNode()`. Chat and Trajectory register separate business Definitions even when they recognize the same durable Event family, while the shared Assembler supplies the same matching, replay, Location, and publication mechanics to both targets.

 #### No generic `end()`

@@ -328,7 +328,9 @@ When business logic deliberately changes a materialized Node to hidden, it leave

 The concrete Tool renderer remains governed by the [`ui-tool ownership decision`](2026-08-08-client-tool-presentation-ownership.md). Tool Definition supplies recursive root/subcall data, and `ui-tool` dispatches concrete presentation by the Tool-name keyed slot.

-Trajectory has no registered target and does not consume the Chat Builder's legacy slice. Its activated `SessionHistoryInspection` keeps an independent history fold, while the ordinary Session snapshot no longer runs a second transcript fold. The Chat Builder retains its legacy slice for StatsLine and the top-level public compatibility fields; a future Trajectory migration does not change the Event Definition, Context, Reader, or Location contracts.
+Trajectory registers its own target and business Definitions against the same Assembler and Session event window as Chat. Its target builder preserves the stage-oriented read model without consuming the Chat Builder's legacy slice or running an independent history fold. The Chat Builder retains its legacy slice for StatsLine and the top-level public compatibility fields; target-specific Definitions do not change the shared Context, Reader, or Location contracts.
+
+The target-specific Trajectory Definitions, retained stage model, Steering adaptation, complexity bounds, and presentation hot paths are owned by the [Trajectory Context assembly decision](2026-08-11-trajectory-conversation-context-assembly.md).

 ## Runtime and render path

@@ -339,20 +341,17 @@ Session Event window
       -> Context matches + State + Location
       -> Definition.buildLocationData(step -> turn)
            -> StepLocation.data / TurnLocation.data
-       -> Definition.buildViewNode(target = chat)
-  -> ChatSnapshotBuilder
-       -> order[] + keyed Node store + Location index + timeline
-  -> ChatView
-       -> ChatNodeSeat(key)
-       -> conversation.chat.node(entryKey = node.kind, hookContext = key)
-            -> slot-level useTurnData(businessKey)
+       -> Definition.buildViewNode() for its declared target
+  -> target View Builder
+       -> chat: ChatSnapshotBuilder -> ChatView -> keyed ChatNodeSeat
+       -> trajectory: TrajectorySnapshotBuilder -> stages/layout/table
 ```

 ## Verification

 Runtime tests pin Definition lifecycle registration, exact-ID append, update-before-start collection followed by forward replay after start, prepend identity, Reader window-gap repair, transitive dependencies, Location closure, Step→Turn data phase order, Location data replacement, publication cadence, illegal withdrawal, and per-target Builders.

-Conversation tests cover every built-in Definition, Assistant Step data, Turn Tail and Deliverables Turn data, Chat ordering and structural sharing, selector isolation, Assistant and Tool running-to-settled identity, nested Code Dispatch, steering, Compaction, Retry, interruption, load-older anchoring, and slot dispatch.
+Conversation tests cover every built-in Chat Definition, Assistant Step data, Turn Tail and Deliverables Turn data, Chat ordering and structural sharing, selector isolation, Assistant and Tool running-to-settled identity, nested Code Dispatch, steering, Compaction, Retry, interruption, load-older anchoring, and slot dispatch. Trajectory tests cover its independently registered Message, Assistant, Tool, Compaction, Request-header, and boundary Definitions together with the preserved stage-oriented view model.

 Slot type/runtime tests pin required parent-provided common inject, the `hookContext` type, Hook isolation across Node contexts, stable factory/Hook identity, and the absence of business-renderer rerenders for unrelated Session publications. Existing entry-owned Observable Hook tests continue to pin the path that does not use a contextual factory.

@@ -382,7 +381,7 @@ History-path tests cover complete replace, non-overlapping prepend, overlapping-

 **Add generic `end()`, prepared, or window-reset lifecycles.** Rejected: businesses have different completion conditions, and a pagination gap is not a business lifecycle. Business Events update State, Location close triggers replay/build, and Reader dependencies own pagination invalidation.

-**Register separate Event Definitions for Chat and Trajectory.** Rejected: identity, State, and Location are target-neutral. `buildViewNode(target)` and each Builder express view differences; Trajectory's independent history fold remains until it registers its own Builder.
+**Reuse one Event Definition across Chat and Trajectory by branching in `buildViewNode(target)`.** Rejected: the views require different business State and intermediate records, so a shared Definition would make each package carry the other's conditions and payloads. Separate target-owned Definitions keep those choices local while sharing the Assembler's ingestion and lifecycle contracts.

 **Add a generic layout model above final business Nodes.** Rejected: activity, tail candidacy, and layout enums would centralize current Chat business semantics in the engine again. Final Nodes carry renderer-required data directly and share only identity, ordering, and Location facts.

@@ -406,4 +405,4 @@ Steps and Turns become stable homes for cross-business aggregates. Turn Tail and

 The cost is new Runtime contracts for Registry, Assembler, Location data, dependency replay, and per-target Builders, plus parent-owned common inject and per-occurrence `hookContext` in UI Slots. Definition authors must understand stable IDs, unique starts, forward replay, Step→Turn publication order, read-only Reader access, and the prohibition on Node withdrawal.

-`useTurnData()` does not revoke the standard `useSession` capability from session-scoped renderers, so this boundary relies on API guidance and tests rather than capability isolation. Registry changes remain low-frequency full rebuilds; the Chat Builder still maintains a legacy slice for StatsLine and the top-level public fields, Trajectory still owns an independent history fold, and built-in Definitions currently remain centralized in `ui-conversation`. These compatibility boundaries do not return business interpretation to Session.
+`useTurnData()` does not revoke the standard `useSession` capability from session-scoped renderers, so this boundary relies on API guidance and tests rather than capability isolation. Registry changes remain low-frequency full rebuilds; the Chat Builder still maintains a legacy slice for StatsLine and the top-level public fields, while Trajectory owns target-specific Definitions and a Builder over the shared Session window. Built-in Definitions remain in their respective UI packages, and these compatibility boundaries do not return business interpretation to Session.
@@ -148,7 +148,7 @@ Assembler 校验 Node `key === context.key` 且 Node `target === target`。业

 `current` 让 Definition 区分“从未生成”与“已经生成后需要隐藏”。Assistant retry 和 Turn Error suppression 使用它避免非法的 Node 撤回。

-Definition 可以针对 target 分支构造不同 data，但匹配、Context identity 和 State 保持 target-neutral。本次只注册 `chat` builder；在拥有注册 target 之前，Trajectory 继续使用独立的 `session-history` fold。
+一个 Definition 最多拥有一个 view target；仅维护状态的 Definition 同时省略 `target` 与 `buildViewNode()`。即使 Chat 与 Trajectory 识别同一持久 Event 族，它们也分别注册自己的业务 Definition；共享 Assembler 则为两个 target 提供相同的匹配、replay、Location 与发布机制。

 #### 不提供通用 `end()`

@@ -328,7 +328,9 @@ Assistant streaming 到 final、Tool running 到 settled 只更新同一个 Seat

 具体 Tool renderer 仍由 [`ui-tool ownership decision`](2026-08-08-client-tool-presentation-ownership.md) 约束。Tool Definition 只交付递归 root/subcall data，`ui-tool` 再按 Tool name keyed slot 分发具体表现。

-Trajectory 尚未注册 target，也不消费 Chat Builder 的 legacy slice。它已激活的 `SessionHistoryInspection` 继续维护独立 history fold，而普通 Session snapshot 不再运行第二套 transcript fold。Chat Builder 为 StatsLine 和顶层公共兼容字段保留 legacy slice；未来迁移 Trajectory 不改变 Event Definition、Context、Reader 或 Location 契约。
+Trajectory 针对与 Chat 相同的 Assembler 和 Session 事件窗口注册自己的 target 与业务 Definition。它的 target builder 保留 stage-oriented read model，既不消费 Chat Builder 的 legacy slice，也不运行独立 history fold。Chat Builder 为 StatsLine 和顶层公共兼容字段保留 legacy slice；target 专属 Definition 不改变共享的 Context、Reader 或 Location 契约。
+
+target 专属 Trajectory Definition、保留的 stage model、Steering 适配、复杂度上界与表现层热点由 [Trajectory Context 组装决策](2026-08-11-trajectory-conversation-context-assembly.md)负责。

 ## Runtime and render path

@@ -339,20 +341,17 @@ Session Event window
       -> Context matches + State + Location
       -> Definition.buildLocationData(step -> turn)
            -> StepLocation.data / TurnLocation.data
-       -> Definition.buildViewNode(target = chat)
-  -> ChatSnapshotBuilder
-       -> order[] + keyed Node store + Location index + timeline
-  -> ChatView
-       -> ChatNodeSeat(key)
-       -> conversation.chat.node(entryKey = node.kind, hookContext = key)
-            -> slot-level useTurnData(businessKey)
+       -> Definition.buildViewNode() for its declared target
+  -> target View Builder
+       -> chat: ChatSnapshotBuilder -> ChatView -> keyed ChatNodeSeat
+       -> trajectory: TrajectorySnapshotBuilder -> stages/layout/table
 ```

 ## Verification

 Runtime tests 固定 Definition 生命周期注册、exact-ID append、update-before-start 收集与 start 后正序 replay、prepend identity、Reader window-gap 修复、传递依赖、Location closure、Step→Turn data phase order、Location data replacement、publication cadence、非法撤回和 per-target Builder。

-Conversation tests 覆盖全部内建 Definition、Assistant Step data、Turn Tail 与 Deliverables Turn data、Chat 排序和结构共享、selector isolation、Assistant/Tool running-to-settled identity、nested Code Dispatch、steering、Compaction、Retry、interruption、load-older anchoring 和 slot dispatch。
+Conversation tests 覆盖全部内建 Chat Definition、Assistant Step data、Turn Tail 与 Deliverables Turn data、Chat 排序和结构共享、selector isolation、Assistant/Tool running-to-settled identity、nested Code Dispatch、steering、Compaction、Retry、interruption、load-older anchoring 和 slot dispatch。Trajectory tests 则覆盖它独立注册的 Message、Assistant、Tool、Compaction、Request-header 与 boundary Definition，以及继续保留的 stage-oriented view model。

 Slot type/runtime tests 固定父注册必须提供声明的 common inject、`hookContext` 类型、不同 Node context 的 Hook 隔离、factory/Hook identity 稳定，以及无关 Session publication 不重渲染业务 renderer。原 entry-owned Observable Hook 测试继续固定未使用 contextual factory 的路径。

@@ -382,7 +381,7 @@ Assembled Web snapshot、GUI 和浏览器场景覆盖真实 plugin graph。浏

 **增加通用 `end()`、prepared 或 window reset 生命周期。** 拒绝：不同业务完成条件不同，分页缺口也不是业务生命周期。业务 Event 更新 State，Location close 触发 replay/build，Reader dependency 负责补页失效。

-**为 Chat 与 Trajectory 注册两套 Event Definition。** 拒绝：identity、State 和 Location 与 target 无关。视图差异由 `buildViewNode(target)` 和各自 Builder 表达；Trajectory 在注册自己的 Builder 之前继续使用独立 history fold。
+**在同一个 Event Definition 内通过 `buildViewNode(target)` 为 Chat 与 Trajectory 分支。** 拒绝：两种视图需要不同的业务 State 与中间记录，共用 Definition 会迫使每个 package 携带另一边的条件与 payload。target 自有的 Definition 把这些选择留在本地，同时复用 Assembler 的摄入与生命周期契约。

 **在最终业务 Node 上再叠一层通用 layout model。** 拒绝：activity、tail candidacy 和 layout enum 会把当前 Chat 的业务语义重新集中到引擎。最终 Node 直接携带 renderer 所需 data，只共享 identity、排序和 Location 事实。

@@ -406,4 +405,4 @@ Step/Turn 成为业务间共享聚合的稳定宿主。Turn Tail 和 Deliverable

 代价是 Runtime 新增 Registry、Assembler、Location data、依赖重放和 per-target Builder 契约，UI Slots 也新增 parent-owned common inject 与 per-occurrence `hookContext`。Definition 作者必须理解稳定 ID、唯一 start、正序 replay、Step→Turn 发布顺序、只读 Reader 和 Node 不撤回规则。

-`useTurnData()` 不撤销 session-scoped renderer 的标准 `useSession`，因此该边界依靠 API 引导和测试，而不是能力隔离。Registry 变化仍是低频完整 rebuild；Chat Builder 继续为 StatsLine 和顶层公共字段维护 legacy slice，Trajectory 继续拥有独立 history fold，内建 Definitions 暂时集中在 `ui-conversation`。这些是兼容边界，不把业务解释权交还给 Session。
+`useTurnData()` 不撤销 session-scoped renderer 的标准 `useSession`，因此该边界依靠 API 引导和测试，而不是能力隔离。Registry 变化仍是低频完整 rebuild；Chat Builder 继续为 StatsLine 和顶层公共字段维护 legacy slice，Trajectory 则在共享 Session 窗口上拥有 target 专属 Definition 与 Builder。内建 Definition 分别留在所属 UI package；这些兼容边界不把业务解释权交还给 Session。
@@ -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/2026-08-09-headless-direct-core-entry-point.md
-2026-08-09-headless-direct-core-entry-point.md: 49afe2993de7302adbedcdf9e8e2347d6424ee2a
-2026-08-09-headless-direct-core-entry-point.zh.md: 73c1cbe5ac777025f63f46751b1d5ccebbfe9676
+2026-08-09-headless-direct-core-entry-point.md: b705df2e6d88e096ee3ba50a6156b815dbd98b98
+2026-08-09-headless-direct-core-entry-point.zh.md: 439f4c21a2ce1741e8d483bc307508550da1bec7
@@ -20,11 +20,11 @@ The shipped `headless` profile contains `dsh-base` and `dsh-headless`. The headl

 `loadProfile` recognizes the exact installation-owned headless tuple (`dsh-base`, `dsh-web-app`, `dsh-headless`) and normalizes it to the shipped headless template while preserving every other manifest field. Extra, missing, or reordered bundle lists are user-owned and remain untouched.

-This note owns the headless transport and completion contracts. [`dsh run` owns one-shot headless execution](../feature/2026-08-08-dsh-run-headless-command.md) owns the command grammar, [GUI layering and RPC protocol](2026-07-19-gui-layering-and-rpc-protocol.md) owns browser gateway boundaries, [web config-tree boot and transport layering](2026-07-24-web-config-tree-boot-and-transport-layering.md) owns the Web tree, and [the default model follows the picker](../feature/2026-08-07-default-model-follows-the-picker.md) owns persistence of the shared Agent default.
+This note owns the headless transport and completion contracts. [Apps own their command lines](2026-08-06-app-owned-command-line.md) owns the current `dsh --profile headless` grammar; the former [`dsh run` decision](../../archived/feature/2026-08-08-dsh-run-headless-command.md) records the superseded launcher-owned grammar, [GUI layering and RPC protocol](2026-07-19-gui-layering-and-rpc-protocol.md) owns browser gateway boundaries, [web config-tree boot and transport layering](2026-07-24-web-config-tree-boot-and-transport-layering.md) owns the Web tree, and [the default model follows the picker](../feature/2026-08-07-default-model-follows-the-picker.md) owns persistence of the shared Agent default.

 ## Verification

-Package tests use the real Session store and Agent registry around a scripted Agent factory to pin idle-to-idle aggregation, late asynchronous completion, terminal model diagnostics, other non-completed exits, direct failures, Loader-time disposal, and flush-before-exit ordering. The keyless assembled snapshots drive `dsh run` through a replayed tool round trip, record a `user/message` with `source.kind: 'user'`, and expose a terminal model failure on stderr. Built-bin acceptance reaches a mock provider through the published entry and requires final text on stdout, exit 0, and empty stderr. Config-dump acceptance excludes every Host, Web, and Client package from the shipped headless tree; PTY shutdown coverage requires no observation line and bounded disposal.
+Package tests use the real Session store and Agent registry around a scripted Agent factory to pin idle-to-idle aggregation, late asynchronous completion, terminal model diagnostics, other non-completed exits, direct failures, Loader-time disposal, and flush-before-exit ordering. The keyless assembled snapshots drive `dsh --profile headless` through a replayed tool round trip, record a `user/message` with `source.kind: 'user'`, and expose a terminal model failure on stderr. Built-bin acceptance reaches a mock provider through the published entry and requires final text on stdout, exit 0, and empty stderr. Config-dump acceptance excludes every Host, Web, and Client package from the shipped headless tree; PTY shutdown coverage requires no observation line and bounded disposal.

 ## Alternatives considered

@@ -39,6 +39,6 @@ Package tests use the real Session store and Agent registry around a scripted Ag

 ## Consequences

-`dsh run` provides a local Agent task rather than browser observation, Host APIs, or HTTP. Users who need those capabilities choose `dsh web`. Successful stderr is empty, completion follows durable flush, and the persisted Session remains available to later tooling. Its initial user message records `source.kind: 'user'` and therefore carries no ApiProxy `rpcId`.
+`dsh --profile headless` provides a local Agent task rather than browser observation, Host APIs, or HTTP. Users who need those capabilities choose `dsh web`. Successful stderr is empty, completion follows durable flush, and the persisted Session remains available to later tooling. Its initial user message records `source.kind: 'user'` and therefore carries no ApiProxy `rpcId`.

 ApiProxy carrier coverage stays in the ApiProxy package. Custom one-shot profiles may include Host or Web bundles explicitly, while the shipped profile and the recognized installation-owned tuple are Web-free.
@@ -20,11 +20,11 @@ Status: implemented

 `loadProfile` 识别安装过程拥有的精确 headless 元组（`dsh-base`、`dsh-web-app`、`dsh-headless`），将其规范化为随附的 headless 模板，并保留 manifest（元数据清单）的其他所有字段。带额外项、缺少项或顺序不同的组合包列表归用户所有，保持不变。

-本 Agent Note 负责 headless 的传输与完成约定。[`dsh run` 负责一次性 headless 执行](../feature/2026-08-08-dsh-run-headless-command.md)负责命令语法，[GUI 分层与 RPC 协议](2026-07-19-gui-layering-and-rpc-protocol.md)负责浏览器网关边界，[Web 配置树启动与传输分层](2026-07-24-web-config-tree-boot-and-transport-layering.md)负责 Web 插件树，[默认模型跟随选择器](../feature/2026-08-07-default-model-follows-the-picker.md)负责共享 Agent 默认值的持久化。
+本 Agent Note 负责 headless 的传输与完成约定。[应用持有自己的命令行](2026-08-06-app-owned-command-line.md)负责当前的 `dsh --profile headless` 语法；原 [`dsh run` 决策](../../archived/feature/2026-08-08-dsh-run-headless-command.md)记录已被取代的启动器持有语法，[GUI 分层与 RPC 协议](2026-07-19-gui-layering-and-rpc-protocol.md)负责浏览器网关边界，[Web 配置树启动与传输分层](2026-07-24-web-config-tree-boot-and-transport-layering.md)负责 Web 插件树，[默认模型跟随选择器](../feature/2026-08-07-default-model-follows-the-picker.md)负责共享 Agent 默认值的持久化。

 ## 验证

-包测试围绕脚本化 Agent 工厂使用真实的会话存储与 Agent 注册表，固定空闲态到空闲态的聚合、延迟异步完成、终止态模型诊断、其他未完成退出、直接失败、Loader 加载期间的 dispose（资源释放），以及退出前 flush 的顺序。组装后的无密钥快照通过回放的工具往返驱动 `dsh run`，记录一条带 `source.kind: 'user'` 的 `user/message`，并在 stderr 暴露终止态模型失败。构建后二进制验收通过已发布入口访问 mock 提供方，并要求最终文本出现在 stdout、退出状态为 0 且 stderr 为空。配置转储验收排除随附 headless 树中的所有 Host、Web 与 Client 包；PTY 关闭覆盖要求不出现观察行，并在有界时间内完成 dispose。
+包测试围绕脚本化 Agent 工厂使用真实的会话存储与 Agent 注册表，固定空闲态到空闲态的聚合、延迟异步完成、终止态模型诊断、其他未完成退出、直接失败、Loader 加载期间的 dispose（资源释放），以及退出前 flush 的顺序。组装后的无密钥快照通过回放的工具往返驱动 `dsh --profile headless`，记录一条带 `source.kind: 'user'` 的 `user/message`，并在 stderr 暴露终止态模型失败。构建后二进制验收通过已发布入口访问 mock 提供方，并要求最终文本出现在 stdout、退出状态为 0 且 stderr 为空。配置转储验收排除随附 headless 树中的所有 Host、Web 与 Client 包；PTY 关闭覆盖要求不出现观察行，并在有界时间内完成 dispose。

 ## 考虑过的替代方案

@@ -39,6 +39,6 @@ Status: implemented

 ## 后果

-`dsh run` 提供本地 Agent 任务，而不是浏览器观察、Host API 或 HTTP。需要这些能力的用户选择 `dsh web`。成功时 stderr 为空，完成结果在持久化 flush 后推导，持久化会话仍可供后续工具使用。初始用户消息记录 `source.kind: 'user'`，因此不携带 ApiProxy `rpcId`。
+`dsh --profile headless` 提供本地 Agent 任务，而不是浏览器观察、Host API 或 HTTP。需要这些能力的用户选择 `dsh web`。成功时 stderr 为空，完成结果在持久化 flush 后推导，持久化会话仍可供后续工具使用。初始用户消息记录 `source.kind: 'user'`，因此不携带 ApiProxy `rpcId`。

 ApiProxy 载体覆盖保留在 ApiProxy 包中。自定义一次性 profile 可以显式包含 Host 或 Web 组合包；随附 profile 与可识别的安装过程所属元组均不含 Web。
@@ -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/2026-08-09-layered-skill-registry.md
-2026-08-09-layered-skill-registry.md: 3f092cfb4b722e3dd51fa4dc46c620259eaffa39
-2026-08-09-layered-skill-registry.zh.md: 38b17329c8d46ee9bbd0863f3fae7cf6be39aa75
+2026-08-09-layered-skill-registry.md: 73897c3cb7e0055ff59221b7ea47c5d6ced06991
+2026-08-09-layered-skill-registry.zh.md: 655780d4ef154434d6debf478134d3293d6c564f
@@ -24,7 +24,7 @@ The composition moves with it: the web-app bundle re-enables the base `skill` re

 **A deployment-level skill reaches every preset-composed session that mounts `tool-skill`.** The repository-plugin e2e's skill root and assertions are restored; the shipped-Web e2e proves the badge row (the same host-registration shape) merges into a standard-preset agent's catalog while the host view stays global-only.

-**Layer visibility and consumption stay separate choices.** A core-web agent can read the global layer in principle, but composes no `skill` tool — whether an agent has skills at all remains the preset's decision, made by mounting or omitting `tool-skill`.
+**Layer visibility and consumption stay separate choices.** A `minimal` agent can read the global layer in principle, but composes no `skill` tool — whether an agent has skills at all remains the preset's decision, made by mounting or omitting `tool-skill`.

 **Provider options are still the borrowed caller object.** `SkillViewOptions` extends `SkillLookupOptions`; the registry consumes `scope` and providers read only their own contract from the same readonly object, preserving the existing borrow-identity guarantee.

@@ -24,7 +24,7 @@ agent-preset stack 曾把整个 skill 能力——注册表、本地提供方和

 **部署级 skill 会到达每个挂载 `tool-skill` 的 preset 会话。**repository-plugin e2e 的 skill 根目录与断言已恢复；shipped-Web e2e 证明 badge 行（同一种宿主注册形态）汇入 standard preset agent 的目录，而宿主视图保持仅全局。

-**层可见性与消费仍是两个独立选择。**core-web agent 原则上可读全局层，但不组合 `skill` 工具——agent 是否拥有 skill 依旧由 preset 通过挂载或省略 `tool-skill` 决定。
+**层可见性与消费仍是两个独立选择。** `minimal` agent 原则上可读全局层，但不组合 `skill` 工具——agent 是否拥有 skill 依旧由 preset 通过挂载或省略 `tool-skill` 决定。

 **提供方选项仍是借用的调用方对象。**`SkillViewOptions` 扩展 `SkillLookupOptions`；注册表消费 `scope`，提供方只从同一个只读对象中读取自己的契约，保持既有的借用恒等保证。

@@ -1,6 +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/2026-07-28-experimental-plugin-package-group.md
-2026-07-28-experimental-plugin-package-group.md: 1ebae5dbb16d4c966f94ffde69fb0cb9bc163d80
-2026-07-28-experimental-plugin-package-group.zh.md: 2d09451c5069a775906e5bc8748c334c29008164
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-08-10-host-plane-ownership-after-presets.md
+2026-08-10-host-plane-ownership-after-presets.md: 5b0a340e875005182a0e6cd0f880b34b14258fb2
+2026-08-10-host-plane-ownership-after-presets.zh.md: 4b1e04f924e656b0f6ad4d070a3b77ce0189c608
@@ -0,0 +1,45 @@
+# Agent Note: What stays host-plane once presets own the agent plane
+
+Status: implemented
+
+English | [中文](2026-08-10-host-plane-ownership-after-presets.zh.md)
+
+## Problem
+
+[Per-session agent presets](2026-08-03-per-session-agent-presets.md) moved every model-facing row onto the agent plane, and each later fix has been one reader that assumed the world before the move. `tasks` came back to the host because a preset row outside its realm resolved it; `goals` never left for the same reason; a child agent's `toolFilter` was repaired once every model-facing tool became an ancestor contribution rather than a global one ([child agents join their parent's preset](../bug-fix/2026-08-10-child-agents-join-their-parent-preset.md)).
+
+Two more readers were still on the wrong side of that line.
+
+`dsh-token-meter` was disabled on the host and mounted inside each preset's `compaction` realm. It takes no configuration, keys every fold by `Session`, and registers no tool or prompt section — but it owns the `tokenUsage`, `contextPressure`, and `contextBreakdown` projection units, and `sessionProjections` is a process-wide table with no scope layering. A unit registered from inside one preset therefore answers for every session: whether a `minimal` session showed a context meter depended on whether some *other* session had mounted `standard` since boot, and a process that only ever ran `minimal` showed none at all.
+
+Nothing named an agent that joined no preset. The join is a scope-parent link; without it the `tools`, `system-prompt`, and `skill` views resolve the empty global layer and the model receives nothing — no error, no empty catalog, just an agent that cannot act. That is how delegated subagents ran for as long as presets existed, and the same hole is open at every entry point that predates them.
+
+## Decision
+
+**The meter is host-plane.** `dsh-token-meter` returns to the host composition and leaves the presets' `isolate` map, so `compact-basic` and `tool-result-prune` resolve the one host instance from inside their realm. The presets keep the realm and the backend — what a preset chooses is whether its agent compacts, not whether its tokens are counted. This is the criterion `tasks` and `goals` are already read by, applied to a Service whose *projection* reach is what made preset ownership wrong: a unit whose empty value is indistinguishable from a real one cannot be per-composition while the table it registers into is per-process.
+
+**An unjoined agent is named twice, at two different points.** `AgentPresets` logs one warning per agent published with a scope chain of length one while a roster is configured. The invariant companion fails instead — and at `system-prompt/assemble`, not at publication, because an unjoined agent is legal until it addresses a model: `recompose` binds exactly such an agent as its first link, and prompt assembly is the only caller that supplies an agent scope, so a host assembly and a standing mount are both correctly out of range.
+
+Three limits stay open and are recorded where they bite rather than fixed here: projection key presence is not a per-session capability signal ([`dsh-session-projection`](../../../../packages/session/session-projection/README.md)); a superseded standing generation is never reclaimed, which the settings-page authoring flow turns into a per-save cost ([`dsh-agent-presets`](../../../../packages/preset/agent-presets/README.md)); and a temporary plugin mounted through `cordis_mount` belongs to the composition rather than the session that mounted it ([`dsh-tool-cordis`](../../../../packages/self-modification/tool-cordis/README.md)).
+
+## Testing
+
+`apps/cli/tests/web-agent-presets.e2e.ts` reads `ctx.get('tokenMeter')` on the booted Web composition before any preset in the file mounts — a preset-side meter sits behind an `isolate` realm and is invisible to `ctx.get`, so the read is an ownership assertion rather than a mount-order coincidence — then asserts a `minimal` session's snapshot carries all three units.
+
+`packages/preset/agent-presets/tests/mount.spec.ts` asserts the warning fires exactly once for a bare agent and not at all for a joined one. `tests/invariant.spec.ts` carries the negative control: an unjoined agent's assembly rejects, while a joined agent's assembly and a scopeless host assembly both pass.
+
+## Alternatives considered
+
+**Keep the meter in the preset and scope-layer the projection registry.** The precise fix, and much larger: `snapshot`, `checkpoint`, and the eager drive would each need a session→scope resolution that a cold read does not have without the api-proxy's `presenterScopeFor`. Rejected as disproportionate to one Service with no per-preset state at all; the general rule is documented on the registry instead.
+
+**Veto publication for an unjoined agent.** Loud beats silent, and the registry supports it — a synchronous `agent/created` listener that throws rolls the creation back. Rejected because composing an agent outside the roster is legal: `recompose` documents the bare agent it then binds, and the ACP bridge, the SDK server, and the headless bundle all create one today. A veto would convert a capability gap into an outage.
+
+**Check the join at `agent/created` in the companion too.** Rejected: publication cannot distinguish a missed join from an agent that will be bound later, so the check would reject a documented path. Prompt assembly can distinguish them.
+
+**Move `plan-mode` and `tool-todo` off the agent plane for the same projection reason.** Rejected: both are genuinely per-preset capabilities, and their units compute an empty value for a session that never uses them, which clients already read by value (`plan.active`, an empty list). Only a unit whose empty value is indistinguishable from a real one — the meter — forces host ownership.
+
+## Consequences
+
+The context meter becomes a per-session fact instead of a function of mount history. A preset can no longer opt out of token accounting; no shipped preset did, and `minimal` now says it drops auto-compaction rather than the accounting.
+
+The warning is advisory, so a deployment that adds a roster to the ACP or SDK-server entry points still starts agents with no tools — it just says so once per agent instead of silently. The invariant reaches only compositions that load `dsh-invariants`, which fences package tests and development hosts, not a shipped one.
@@ -0,0 +1,45 @@
+# Agent Note: What stays host-plane once presets own the agent plane
+
+Status: implemented
+
+[English](2026-08-10-host-plane-ownership-after-presets.md) | 中文
+
+## Problem
+
+[逐会话 agent preset](2026-08-03-per-session-agent-presets.md) 把每一个面向模型的行搬上了 agent 平面，此后的每一处修复都是一个仍按搬迁之前的世界写成的读取点。`tasks` 因为 realm 之外的 preset 行要解析它而搬回宿主；`goals` 因为同样的理由从未离开；而当所有面向模型的工具都变成祖先贡献之后，子 agent 的 `toolFilter` 也已被修好（[子 agent 加入父方 preset](../bug-fix/2026-08-10-child-agents-join-their-parent-preset.md)）。
+
+还有两个读取点仍站在这条线的错误一侧。
+
+`dsh-token-meter` 在宿主侧被禁用，改挂进每个 preset 的 `compaction` realm。它不接受任何配置，每次折叠都以 `Session` 建键，也不注册工具或提示段——但它拥有 `tokenUsage`、`contextPressure` 与 `contextBreakdown` 三个投影单元，而 `sessionProjections` 是一张进程级、没有作用域分层的表。因此从某个 preset 内部注册的单元会替所有会话作答：一个 `minimal` 会话是否显示 context meter，取决于本次启动以来有没有**别的**会话挂过 `standard`；而只跑过 `minimal` 的进程根本不显示。
+
+没有加入任何 preset 的 agent 也无人指出。加入是一条 scope 父链链接；缺了它，`tools`、`system-prompt` 与 `skill` 的视图都解析到空的全局层，模型什么也收不到——不报错，也没有空目录可看，只是一个无法行动的 agent。被委派的子 agent 在 preset 存在的整段时间里都是这样运行的，而同一个洞在每一个早于 preset 的入口点上都开着。
+
+## Decision
+
+**meter 属于宿主平面。** `dsh-token-meter` 回到宿主组装，并离开各 preset 的 `isolate` 映射，于是 `compact-basic` 与 `tool-result-prune` 在自己的 realm 内部解析到那一份宿主实例。preset 保留 realm 与压缩后端——preset 选择的是它的 agent 是否压缩，而不是它的 token 是否被计。这正是 `tasks` 与 `goals` 已经采用的判据，只是这次适用于一个因**投影**触达面而不该归 preset 所有的 Service：当一个单元的空值与真实值无法区分时，只要它注册进的那张表是进程级的，它就不能是逐组装的。
+
+**未加入的 agent 在两个不同的点上被指出两次。** 在配置了名单的前提下，`AgentPresets` 对每个作用域链长度为一就发布的 agent 记录一条警告。invariant 配套则直接失败——并且发生在 `system-prompt/assemble` 而非发布时，因为一个未加入的 agent 在它对模型说话之前都是合法的：`recompose` 绑定的正是这样一个 agent 作为它的首次链接；而提示词组装是唯一会提供 agent 作用域的调用方，因此宿主组装与常驻挂载都正确地落在检查范围之外。
+
+有三处限制不在此处修复，而是记录在会咬到它们的地方：投影 key 是否存在不能当作逐会话的能力信号（[`dsh-session-projection`](../../../../packages/session/session-projection/README.md)）；被替代的常驻代际永不回收，而设置页的编写流程把它变成每次保存的代价（[`dsh-agent-presets`](../../../../packages/preset/agent-presets/README.md)）；通过 `cordis_mount` 挂上的临时插件属于组装而非挂载它的会话（[`dsh-tool-cordis`](../../../../packages/self-modification/tool-cordis/README.md)）。
+
+## Testing
+
+`apps/cli/tests/web-agent-presets.e2e.ts` 在本文件中任何 preset 挂载**之前**，于已启动的 Web 组装上读取 `ctx.get('tokenMeter')`——preset 侧的 meter 会待在 `isolate` realm 里，对 `ctx.get` 不可见，因此这次读取是一次所有权断言而不是挂载顺序的巧合——随后断言一个 `minimal` 会话的快照带齐三个单元。
+
+`packages/preset/agent-presets/tests/mount.spec.ts` 断言警告对裸 agent 恰好触发一次、对已加入的 agent 完全不触发。`tests/invariant.spec.ts` 承担负控：未加入 agent 的组装被拒绝，而已加入 agent 的组装与不带作用域的宿主组装都通过。
+
+## Alternatives considered
+
+**把 meter 留在 preset，改为给投影注册表分层。** 这是更精确的修法，代价也大得多：`snapshot`、`checkpoint` 与主动驱动都需要一次「会话 → 作用域」的解析，而冷读在没有 api-proxy 的 `presenterScopeFor` 时并不具备。相对于一个完全没有 per-preset 状态的 Service，这不成比例，因此改为把通则写在注册表上。
+
+**对未加入的 agent 否决发布。** 大声胜过静默，注册表也支持这么做——同步的 `agent/created` 监听器抛出会把创建整体回滚。否决的理由是：在名单之外组装 agent 是合法的——`recompose` 写明了它随后绑定的那个裸 agent，而 ACP 桥、SDK server 与 headless bundle 今天都会创建一个。否决会把能力缺口变成一次故障。
+
+**让配套也在 `agent/created` 处检查加入情况。** 否决：发布时分不清漏掉的加入与之后才会被绑定的 agent，因此该检查会拒绝一条已写明的路径。提示词组装分得清。
+
+**基于同样的投影理由，把 `plan-mode` 与 `tool-todo` 也搬离 agent 平面。** 否决：两者确实是逐 preset 的能力，且对从不使用它们的会话，其单元算出的就是空值，而客户端本来就按值读取（`plan.active`、空列表）。只有空值与真实值无法区分的单元——meter——才被迫归宿主所有。
+
+## Consequences
+
+context meter 成为逐会话的事实，而不再是挂载历史的函数。代价是 preset 不能再选择不做 token 记账；随附的 preset 没有一个这么做，`minimal` 现在也写明它放弃的是自动压缩而非记账。
+
+那条警告是建议性的，因此给 ACP 或 SDK server 入口加上名单的部署依然会启动没有工具的 agent——只是每个 agent 会说一次，而不再静默。invariant 只触达装载了 `dsh-invariants` 的组装，因此它把关的是包测试与开发宿主，不是随附宿主。
@@ -1,6 +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/2026-07-30-config-only-repository-plugins.md
-2026-07-30-config-only-repository-plugins.md: 35327a30e03c51311f634e05ade209ab93ae0155
-2026-07-30-config-only-repository-plugins.zh.md: 5755045560da761b59f7c65e99d551f599c2b5b3
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-08-10-message-feedback-sidecar.md
+2026-08-10-message-feedback-sidecar.md: 780cbaa840fcac7bcfa799468bfd61f5b08715cb
+2026-08-10-message-feedback-sidecar.zh.md: 72ecc82717010f65d013418a45c3b38c6084aaf9
@@ -0,0 +1,43 @@
+# Agent Note: Lifecycle-bound message feedback sidecar
+
+Status: implemented
+
+English | [中文](2026-08-10-message-feedback-sidecar.zh.md)
+
+## Problem
+
+The existing `/feedback` command records an immutable Session-level `feedback/record` event. That event can release a pending telemetry prefix under `FEEDBACK_ONLY`, so it is the wrong authority for an editable positive/negative rating and optional note attached to one assistant message. Message feedback needs independent update and delete semantics without entering the canonical Session log, changing a projection, reaching the model surface, or implicitly consenting to telemetry.
+
+A sidecar keyed only by `SessionId` can outlive the log lifecycle it describes when an id is recreated with a different header identity. A Session-wide revision also makes unrelated message edits conflict, while plain storage-domain read/put has no cross-process compare-and-swap. Session disposal is only live-store detach, not durable deletion, and the current Session persistence seam exposes no deletion operation that could own a truthful cascade.
+
+## Decision
+
+`@deepseek-ai/dsh-message-feedback` owns the `ctx.messageFeedback` service and stores message feedback as one storage-domain sidecar row per Session. The sidecar is neither Session-log content nor a Session projection. It emits no `feedback/record` event and performs no telemetry handoff; the command-feedback and message-feedback contracts remain independent.
+
+Every usable row is bound to the inspected Session header identity `{createdAt, cwd}`, not merely its `SessionId`. A lifecycle mismatch is treated as absence: `list` returns no items, and `put` may replace the stale row with one bound to the current identity. An id reused with a different header identity therefore cannot inherit stale feedback. A fork receives its own Session identity and no sidecar copy: even when the fork seed contains the same assistant messages, feedback remains attached to the Session in which the human recorded it.
+
+`put` accepts a target only when `SessionPersistence.inspect()` observes a non-empty, append-origin `assistant/message` with that `MessageId`. Replacement-origin messages, empty usage-only assistant records, and non-assistant targets are rejected. Inspection is the cold-safe authority: it neither publishes or resumes an Agent nor commits cold-log repair merely to validate feedback. A cold `listSnapshots()` preflight classifies definite absence; inspection failure for a catalogued Session remains an infrastructure failure. A request in the narrow live-detach-to-header-materialization interval can therefore return `session-not-found`, and the caller retries after retirement materialization.
+
+Before `put` commits a sidecar row, it puts the target log behind a durability barrier. A matching live Session passes through the canonical `ctx.sessions.flush` checkpoint, then both live and cold paths are physically read from sequence zero through `SessionPersistence.readFrom`. The resulting observation's header identity and target are checked again. A missing flush participant, changed identity, vanished target, or physical-read failure prevents the sidecar write, so a committed feedback item never precedes the durable assistant message it references.
+
+Each message item carries its own opaque version plus Host-assigned `createdAt` and `updatedAt` timestamps. `put` compares the caller's `ifVersion` only with the addressed item, so editing one message does not invalidate another. The comparison is strict even when the desired value already matches, preventing a stale request from crossing an ABA value cycle; a conflict returns the authoritative current item so callers can reconcile without a second read. A matching-version no-op preserves the version and timestamps, while a material update preserves `createdAt`, replaces the version, and keeps `updatedAt` from moving backward. An already-absent delete is likewise successful. Versions are tokens for equality, not counters callers may order or synthesize.
+
+A per-Session mutation queue encloses lifecycle inspection, sidecar read, conflict evaluation, and whole-row write. This makes one service instance's mutations serial and preserves the per-message compare-and-swap contract inside one Host process. Plugin disposal closes admission, drains accepted queue work, and then closes the storage domain. The underlying storage-domain API provides no cross-process conditional write, so the implementation claims no cross-process linearizability or lost-update protection.
+
+`maxNoteBytes` is a required deployment choice and bounds the UTF-8 byte length of an optional note; the Web Host bundle sets it explicitly to `8192`. The package publishes the Host `messageFeedback.list`, `messageFeedback.put`, and `messageFeedback.delete` contract directly through `GatewayService` and `@Remote`. Client Remote aggregate mounting and UI remain separately owned and deferred; their later adapter stays a thin consumer of this Host contract.
+
+The service performs no fake deletion cascade. `session/disposed` and `host/session-removed` describe detach from live ownership, not durable Session deletion, and Session persistence currently has no delete surface. Sidecar rows can therefore remain after out-of-band log removal; a different `{createdAt, cwd}` prevents such an orphan from becoming feedback for a later Session that reuses the id.
+
+## Alternatives considered
+
+**Append edits to the Session log and derive a projection.** Rejected because editable UI metadata would become canonical conversation-adjacent history, forks would replay and inherit it, deletion would require tombstones, and reusing `feedback/record` would silently couple a message rating to telemetry consent.
+
+**Key feedback globally by `MessageId`, copy it on fork, or use one Session revision.** Rejected because message ids are meaningful only within a Session lifecycle, forked conversations need independent human judgments, and unrelated message mutations must not create false conflicts.
+
+**Extend `KvTable` with cross-process compare-and-swap in this change.** Rejected because the shipped storage-domain backends expose no common conditional-write primitive. A process-local queue matches the supported one-Host topology; a real multi-process guarantee requires a backend-level atomic contract and is separate work.
+
+**Delete feedback on Session disposal.** Rejected because disposal includes ordinary detach and rollback paths. Treating it as durable deletion would lose feedback while the Session log still exists; cleanup waits for a real Session deletion authority.
+
+## Consequences
+
+Message feedback is locally durable and independently editable without changing model-visible history or telemetry behavior. Concurrent callers in one Host receive per-message conflict detection and retry-safe outcomes, while deployments with multiple writers to the same storage root remain unsupported. A differing header identity treats a stale row as absent but does not reclaim it; a cloned log that retains the same `{createdAt, cwd}` is indistinguishable by this contract. The Host Remote contract is available now; client assembly and UI can remain thin consumers rather than taking ownership of persistence or concurrency semantics.
@@ -0,0 +1,43 @@
+# Agent Note: 绑定生命周期的消息反馈伴随记录
+
+Status: implemented
+
+[English](2026-08-10-message-feedback-sidecar.md) | 中文
+
+## 问题
+
+现有 `/feedback` 命令记录不可变的 Session 级 `feedback/record` 事件。在 `FEEDBACK_ONLY` 下，该事件可以释放待处理的遥测前缀，因此它不适合作为挂在单条 assistant 消息上的可编辑好评／差评与可选备注的权威来源。消息反馈需要独立的更新与删除语义，且不得进入权威 Session 日志、改变投影、到达模型接口，或隐式表示遥测同意。
+
+只按 `SessionId` 建索引的伴随记录可能在该 id 以不同 header 身份重建后，继续存活于其所描述的日志生命周期之外。Session 级 revision 还会让无关消息的编辑彼此冲突，而普通 storage-domain 读／写不提供跨进程 compare-and-swap。Session disposal 只是从 live store 脱离，并非持久删除；当前 Session 持久化 seam 也没有可拥有真实级联的删除操作。
+
+## 决策
+
+`@deepseek-ai/dsh-message-feedback` 拥有 `ctx.messageFeedback` 服务，并把消息反馈存为每个 Session 一条 storage-domain 伴随记录（sidecar）。该伴随记录既不是 Session 日志内容，也不是 Session 投影。它不发出 `feedback/record` 事件，也不执行遥测交接；command-feedback 与 message-feedback 契约保持独立。
+
+每条可用记录都绑定到经检查的 Session header 身份 `{createdAt, cwd}`，而不只是其 `SessionId`。生命周期不匹配按不存在处理：`list` 返回空条目，`put` 可以用绑定当前身份的新记录替换陈旧行。因此，以不同 header 身份复用的 id 不会继承陈旧反馈。fork 拥有自己的 Session 身份，且不复制伴随记录：即使 fork 种子包含相同的 assistant 消息，反馈仍只属于人类记录它的那个 Session。
+
+`put` 只接受由 `SessionPersistence.inspect()` 观测到的非空、append-origin `assistant/message`，且其 `MessageId` 必须与目标相同。replacement-origin 消息、仅承载 usage 的空 assistant 记录以及非 assistant 目标都会被拒绝。检查使用 cold-safe 权威路径：它不会仅为验证反馈而发布或恢复 Agent，也不会提交 cold 日志修复。cold 路径由 `listSnapshots()` 预检明确不存在；已进入目录的 Session 若检查失败，仍按基础设施故障处理。因此，请求若恰落在 live detach 到 header materialization 的极短窗口，可能返回 `session-not-found`，调用方在 retirement materialization 后重试。
+
+`put` 提交伴随记录前，会先让目标日志通过 durability barrier。身份匹配的 live Session 经过权威 `ctx.sessions.flush` checkpoint，随后 live 与 cold 路径都会通过 `SessionPersistence.readFrom` 从序列零做物理复读。之后再次校验所得观测的 header 身份与目标。缺少 flush 参与方、身份变化、目标消失或物理读取失败都会阻止伴随记录写入，因此已提交反馈绝不会先于它引用的持久 assistant 消息。
+
+每个消息条目都携带自己的 opaque version，以及 Host 分配的 `createdAt` 和 `updatedAt` 时间戳。`put` 只把调用方的 `ifVersion` 与目标条目比较，因此编辑一条消息不会使另一条消息失效。即使目标值已经相同，比较仍然严格执行，从而防止陈旧请求穿过 ABA 值循环；冲突会返回权威当前条目，调用方无需二次读取即可协调。携带匹配 version 的无变化请求会保留 version 与时间戳；实质更新保留 `createdAt`、替换 version，并保证 `updatedAt` 不倒退。删除已经不存在的条目也同样成功。version 是只能做相等比较的 token，不是调用方可以排序或自行合成的计数器。
+
+按 Session 划分的变更队列覆盖生命周期检查、伴随记录读取、冲突判断与整行写入。这使同一个服务实例的变更串行化，并在单个 Host 进程内保持逐消息 compare-and-swap 契约。Plugin disposal 会关闭接纳、排空已进入队列的工作，然后关闭 storage domain。底层 storage-domain API 不提供跨进程条件写，因此实现不承诺跨进程线性一致性或防止丢失更新。
+
+`maxNoteBytes` 是必填的部署选择，用于限制可选备注的 UTF-8 字节长度；Web Host bundle 将其显式设为 `8192`。该包通过 `GatewayService` 与 `@Remote` 直接发布 Host `messageFeedback.list`、`messageFeedback.put` 与 `messageFeedback.delete` 契约。客户端 Remote 聚合挂载与 UI 由各自边界负责并保持延后；后续适配层只是该 Host 契约的薄消费者。
+
+服务不伪造删除级联。`session/disposed` 与 `host/session-removed` 表示脱离 live ownership，而非持久删除，Session persistence 当前也没有删除接口。因此在带外移除日志后，伴随记录可能继续存在；不同的 `{createdAt, cwd}` 可阻止此类孤儿记录变成后来复用该 id 的 Session 反馈。
+
+## 考虑过的替代方案
+
+**把编辑追加到 Session 日志并派生投影。** 不予采纳，因为可编辑 UI 元数据会变成权威且邻近对话的历史，fork 会回放并继承它，删除需要 tombstone，而复用 `feedback/record` 会把消息评分与遥测同意静默耦合。
+
+**按全局 `MessageId` 建索引、在 fork 时复制，或使用一个 Session revision。** 不予采纳，因为消息 id 仅在某个 Session 生命周期内有意义，fork 后的对话需要独立的人类判断，而且无关消息的变更不应制造虚假冲突。
+
+**在本次变更中为 `KvTable` 扩展跨进程 compare-and-swap。** 不予采纳，因为出厂 storage-domain 后端没有共同的条件写原语。进程内队列符合受支持的单 Host 拓扑；真实的多进程保证需要后端级原子契约，属于独立工作。
+
+**在 Session disposal 时删除反馈。** 不予采纳，因为 disposal 包含普通 detach 与 rollback 路径。把它当成持久删除会在 Session 日志仍存在时丢失反馈；清理必须等待真正的 Session 删除权威。
+
+## 后果
+
+消息反馈在本地持久化并可独立编辑，且不改变模型可见历史或遥测行为。同一 Host 中的并发调用方获得逐消息冲突检测与可安全重试的结果；多个写入者共享同一存储根目录的部署仍不受支持。不同的 header 身份会让陈旧记录被视为不存在，但不会将其回收；本契约无法区分保留相同 `{createdAt, cwd}` 的克隆日志。Host Remote 契约现在可用；客户端组装与 UI 可以保持为薄消费者，而不接管持久化或并发语义。
@@ -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/2026-08-10-product-subagent-providers-in-shared-host.md
+2026-08-10-product-subagent-providers-in-shared-host.md: 33b6eb6cf7a6c19e9ea71cdb7dc8881e8052ef24
+2026-08-10-product-subagent-providers-in-shared-host.zh.md: fd78c7a3fee4e4ee30d27d87c752e1a23576fd85
@@ -0,0 +1,41 @@
+# Agent Note: Product subagent providers live in the shared profile host
+
+Status: implemented
+
+English | [中文](2026-08-10-product-subagent-providers-in-shared-host.zh.md)
+
+## Problem
+
+The [Codex and Claude Code provider contracts](../feature/2026-08-04-claude-code-and-codex-subagent-backends.md) were first shipped as independently installable packages that a deployment loaded beside the common subagent tool. Agent Presets later became the ordinary owner of one agent's model-visible tools, but a preset cannot safely own these product providers: `ctx.subagents` is a process registry, provider names are unique, and host consumers resolve the same registry across sessions. Requiring a person to edit both a Profile and a Preset would also make a generic preset row incomplete by itself.
+
+The placement decision must preserve two independent facts. Loading a provider must not start or authenticate a product, while enabling a tool must remain per preset so two sessions can expose different products. A global product switch, a provider instance per agent, or pre-enumerated combination presets would each create a second owner for one of those facts.
+
+## Decision
+
+Every shipped Profile loads the fixed `codex` and `claude-code` providers once through the base bundle's host plane. Loading either plugin only registers a dormant backend; the corresponding Codex or Claude process starts on the first actual delegation call. Agent Presets independently contribute ordinary `dsh-tool-subagent` rows for `subagent_codex` and `subagent_claude_code`, so a preset can expose neither tool, either one, or both without changing the provider registry.
+
+This decision supersedes only the opt-in composition placement recorded by the provider-contract note. That note continues to own each product protocol, result mapping, cancellation, process-tree lifecycle, and evidence tiers. The [Agent Preset architecture](2026-08-03-per-session-agent-presets.md) continues to own the Host/Agent split, preset authoring, and the rule that edits affect only newly composed sessions.
+
+The providers use products already selected by the host environment. Codex starts `codex` from `PATH`; Claude Code resolves `claude` through the shared subprocess execution world and passes the exact path to the official SDK. Profile loading does not install a product, create product state, probe a version, test authentication, or add product-specific settings. Missing commands and product failures remain local to the attempted delegation.
+
+The current base dependency closure still includes the Claude Agent SDK's optional platform CLI payload even though production resolves the host `claude`. Removing that unused payload belongs to the separate product installation-closure follow-up; this placement decision neither installs it dynamically nor treats it as the production executable.
+
+## Verification
+
+The base Loader test proves both provider names register exactly once and no product process starts during Profile boot. Real Agent Preset composition covers none, Codex-only, Claude-only, and both tool sets, including generation isolation after an authored preset changes. Keyless ACP snapshots pin the model-visible tool schemas for one and both products, while provider tests separately prove native executable resolution, failure, cancellation, and process-tree quiescence.
+
+## Alternatives considered
+
+**Keep product providers opt-in at the Profile layer.** This preserves a smaller default dependency closure, but a copied or agent-authored Preset row is not usable unless the person also discovers and edits a second composition layer. It leaves the general Preset entry incomplete for these otherwise ordinary tools.
+
+**Store global or per-Profile product enable switches.** A process switch competes with the Preset as owner of model-visible tools and cannot express two sessions using different combinations. Availability and authentication are deployment facts, not another persisted product state.
+
+**Mount a provider inside every Agent Preset.** Provider names belong to a process registry, so the second session would collide with the first. Host consumers also need the registry independently of any one agent's lifetime.
+
+**Ship four product-combination presets.** Four identities duplicate complete compositions to represent two independent tool rows. Ordinary rows already express the full matrix without adding roster or maintenance state.
+
+## Consequences
+
+A user manages both products through the same Agent Preset authoring path as other plugins, and each new session receives exactly the tools its chosen preset contributes. Every Profile carries two dormant provider registrations, so unused products consume package and module-loading footprint but no product process, login, model call, or product home.
+
+The Host registry remains the single provider authority and each Preset remains the single model-tool authority. The trade-off is the current Claude SDK optional-payload installation cost, which stays explicitly deferred rather than being hidden behind another enable state or installer lifecycle.
@@ -0,0 +1,41 @@
+# Agent Note: 产品 subagent 提供方位于共享 profile 宿主
+
+Status: implemented
+
+[English](2026-08-10-product-subagent-providers-in-shared-host.md) | 中文
+
+## 问题
+
+[Codex 与 Claude Code 提供方约定](../feature/2026-08-04-claude-code-and-codex-subagent-backends.md)最初以可独立安装的包交付，由部署环境在通用 subagent 工具旁加载。Agent Preset 后来成为单个 agent（智能体）的模型可见工具的常规责任方，但 preset 不能安全地拥有这些产品提供方：`ctx.subagents` 是进程级注册表，提供方名称唯一，而宿主消费方会跨会话解析同一个注册表。如果要求用户同时编辑 Profile 和 Preset，也会使通用 preset 行本身不完整。
+
+归属决策必须同时保留两个彼此独立的事实：加载提供方不得启动产品，也不得对产品执行身份验证；而工具是否启用仍须按 preset 决定，这样两个会话才能暴露不同的产品。全局产品开关、按 agent 创建提供方实例或预先枚举的组合 preset，都会为其中一个事实另设第二责任方。
+
+## 决策
+
+每个随发行版交付的 Profile 都会通过 base 组合包的宿主平面，把固定的 `codex` 与 `claude-code` 提供方各加载一次。加载任一插件只会注册一个休眠后端；对应的 Codex 或 Claude 进程直到第一次实际委派调用时才启动。Agent Preset 分别通过普通的 `dsh-tool-subagent` 行贡献 `subagent_codex` 与 `subagent_claude_code`，因此一个 preset 可以不暴露任何工具、只暴露其中一个或同时暴露两者，而无需更改提供方注册表。
+
+本决策仅取代提供方约定说明所记录的、原先由用户选择启用的组装位置。该说明仍负责每个产品的协议、结果映射、取消、进程树生命周期与证据层级。[Agent Preset 架构](2026-08-03-per-session-agent-presets.md)仍负责宿主与 agent 的划分、preset 创作，以及改动只影响新组装会话的规则。
+
+这些提供方使用宿主环境已经选定的产品。Codex 启动 `codex`，该命令从 `PATH` 解析；Claude Code 通过共享的子进程执行世界解析 `claude`，并把确切路径交给官方 SDK。加载 Profile 不会安装产品、创建产品状态、探测版本、测试身份验证，也不会新增产品专属设置。命令缺失和产品故障仍局限于发生问题的那次委派。
+
+当前 base 依赖闭包仍包含 Claude Agent SDK 的可选平台 CLI（命令行界面）载荷，尽管生产环境解析的是宿主提供的 `claude`。移除这份未使用载荷属于独立的产品安装闭包后续项；本归属决策既不会动态安装它，也不会将它当作生产可执行文件。
+
+## 验证
+
+base Loader 测试证明两个提供方名称都恰好注册一次，而且 Profile 启动期间不会启动产品进程。真实 Agent Preset 组装覆盖不暴露任何工具、仅暴露 Codex、仅暴露 Claude 和同时暴露两者这四种工具集合，也覆盖自行创作的 preset 发生改动后的代际隔离。无密钥 ACP（Agent Client Protocol）快照固定单个产品与两个产品同时启用时的模型可见工具 schema，提供方测试则另行证明原生可执行文件解析、失败、取消和进程树完全停稳。
+
+## 考虑过的替代方案
+
+**将产品提供方保留为 Profile 层的按需启用项。** 这样可缩小默认依赖闭包，但复制或由 agent 创作的 Preset 行无法直接使用，除非用户还发现并编辑第二个组装层。对于这些本来与其他工具无异的工具，通用 Preset 入口仍不完整。
+
+**存储全局或按 Profile 配置的产品启用开关。** 进程级开关会与 Preset 争夺模型可见工具的责任归属，也无法表示两个会话使用不同组合。可用性与身份验证属于部署事实，并非另一份需要持久化的产品状态。
+
+**在每个 Agent Preset 内挂载一个提供方。** 提供方名称属于进程级注册表，因此第二个会话会与第一个冲突。宿主消费方也需要独立于任何单个 agent 的生命周期使用该注册表。
+
+**交付四个产品组合 preset。** 四个身份会复制完整组装，只为表示两条独立的工具行。普通行已经能表达完整矩阵，无需新增名单或维护状态。
+
+## 后果
+
+用户通过与其他插件相同的 Agent Preset 创作路径管理两个产品，每个新会话只会获得其所选 preset 所贡献的工具。每个 Profile 都携带两个休眠的提供方注册，因此未使用的产品会产生包和模块加载开销，但不会启动产品进程、登录、调用模型或创建产品主目录。
+
+宿主注册表仍是提供方的唯一权威，每个 Preset 仍是模型工具的唯一权威。代价是当前 Claude SDK 可选载荷的安装成本继续被明确延期处理，而不会隐藏在另一种启用状态或安装程序生命周期之后。
@@ -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/2026-08-10-session-log-version-mechanism.md
+2026-08-10-session-log-version-mechanism.md: 25eb1230a254219c827b1d2750dba367b113f9f7
+2026-08-10-session-log-version-mechanism.zh.md: c47670f2de77773c17c9595eff442bf7f1e8ec3e
@@ -0,0 +1,30 @@
+# Agent Note: Session log versioning — one integer, an upgrade chain, and a per-event ignorable marker
+
+Status: implemented
+
+English | [中文](2026-08-10-session-log-version-mechanism.zh.md)
+
+## Problem
+
+Session logs must be upgradable after release, and the runtime that ships first is the floor for every later decision: whatever refusal and degradation behavior is missing from the first released reader can never be added to the copies users already run. Release issue #1901 required at minimum that an old runtime reading a newer session format reports "unsupported" instead of misreading it. The pre-change reader did the opposite on both axes: `assertVersion` rejected any version mismatch with one direction-blind message, and the JSONL decoder passed unknown event types through untouched, so reconstruction silently skipped them — resuming a gutted session with no diagnostic at all.
+
+## Decision
+
+**One monotonic integer, no major/minor split.** Whether a version step is auto-upgradable is a property of that step — expressed by whether its upgrader exists — not something a two-level numbering scheme should promise in advance (you rarely know at design time whether the next change will turn out "major"). This matches the SQLite backend's `SCHEMA_VERSION` precedent.
+
+**The writer decides bumps, not the reader.** A bump is required exactly when an old runtime could no longer handle a new log with full semantic correctness. "Parses without error" is not the bar: silently skipping content that shapes reconstruction is a wrong read. Only structural changes qualify — header shape, event envelope, core event semantics, the surface mechanism (`SurfaceEventType` set, `SurfaceOp` variants). When unsure, bump: a near-identity upgrader is almost free, a missed bump silently corrupts old readers.
+
+**Read rules by direction.** Equal version: read normally. Newer than the reader: refuse, name the direction ("written by a newer harness — upgrade"), and point at the raw log artifact so the user can still see the text (`SessionFormatUnsupportedError`, distinct from `SessionPersistenceCorruptionError` because nothing is damaged). Older than the reader: convert in memory through the chain of n→n+1 upgraders for viewing; persist the converted log only when the session is actually continued (atomic temp-file replace, original kept as backup). A step whose upgrader cannot be written is left empty, which cuts off every version at or below it — those degrade to raw-text viewing.
+
+**A per-event `ignorable` marker covers vocabulary growth, so ordinary event additions never bump the version.** The event vocabulary is decided by which plugins are mounted, which a single version integer cannot describe. A reader meeting an unrecognized event type refuses to interpret the log unless the event carries `ignorable: true` in its envelope. The default is *required*: forgetting the marker over-refuses a resumable session (an inconvenience), while a default of ignorable would make the same mistake silently resume a gutted one (a safety failure). The architecture makes this sound: model-visible content flows only through the three `surfaceOp`-marked surface event types plus the `request/header`/`request/context` folds, so the dangerous unknowns are exactly the non-surface events that change how the rest of the log is read (`session/end-seed` is the existing example).
+
+## Consequences
+
+What shipped in v0 (release 0812): direction-aware refusal with the raw-log path; the unknown-event guard against a generated known-vocabulary list (`KNOWN_SESSION_EVENT_TYPES`, emitted by `gen-persistence-catalog` from every `SessionEventMap` merge and kept fresh by `verify-persistence-catalog`); the `ignorable` envelope field accepted by seed validation, both backends (a dedicated SQLite column, `SCHEMA_VERSION` 15), and the BFF wire schema. The upgrader chain itself is deferred until the first real v0→v1 step exists to test it against; writers do not yet set `ignorable` (no producer needs it), so `Session.append` gains that surface with its first user. Until a registration surface exists, an out-of-repo plugin's events refuse resume under first-party readers — the pre-release stance accepts that, and the refusal is loud rather than silent. The unknown-type guard is read-side only: `appendCore` keeps rejecting retired legacy shapes but does not vocabulary-check new types, because an append-time refusal would stall a live session's durability mid-flight, which costs more than a loud refusal at the log's next load. The JSONL backend additionally refuses a foreign version from the raw header line before validating today's header shape or decoding any event row, so a structurally different future format still reports the upgrade direction instead of "corrupt"; SQLite gates whole-file structure through its own `SCHEMA_VERSION` pragma first.
+
+## Alternatives considered
+
+- **Major/minor versioning** — the "is it convertible" bit lives on each step's upgrader, and pre-committing it into a number shape invites wrong promises.
+- **Default-ignorable unknown events** — inverts the failure mode of a forgotten marker from visible over-refusal into silent corruption.
+- **Auto-migrating on view** — rewriting the artifact on open turns a read into a destructive write: a converter bug corrupts logs at browse time, and a same-directory older runtime loses access because a newer one merely looked.
+- **Per-plugin runtime registration of known event types** — would make the known set composition-dependent, so a leaner same-version composition would refuse logs a fuller one wrote. The generated repo-wide list keeps same-version reads uniform; out-of-repo plugin events are outside it by construction, and a registration surface for them is deferred until such a consumer exists.
@@ -0,0 +1,30 @@
+# Agent Note：Session log 版本机制：单调整数、升级器链、逐事件可忽略标记
+
+Status: implemented
+
+[English](2026-08-10-session-log-version-mechanism.md) | 中文
+
+## 问题
+
+Session log 在发布后必须能升级格式，而最先发布的运行时决定了此后一切的下限：第一个发布版的读取器缺少哪种拒绝和降级行为，用户手里已经装上的副本就永远补不上。发布 issue #1901 的最低要求是老运行时读到新 Session 格式时明确报不支持，而不是读错。改动前的读取器在两个方向上都做反了：`assertVersion` 对任何版本不匹配抛出同一条不区分方向的消息；JSONL 解码器把不认识的事件类型原样放行，重建时静默跳过，恢复出一个内容残缺的会话且没有任何诊断。
+
+## 决定
+
+**一个单调递增的整数，不分大小版本。**某一步能不能自动升级是那一步自己的属性，由它的升级器存在与否表达，不该由两级编号方案提前承诺（设计时很少能预知下一个变更算不算"大"）。这与 SQLite 后端 `SCHEMA_VERSION` 的先例一致。
+
+**升不升版本由写入方决定，与读取方能力无关。**当且仅当老运行时无法在语义上完全正确地处理新日志时才必须升版本。"解析不报错"不是标准：静默跳过影响重建的内容就是读错。只有结构性变更够得上这条线：header 形状、事件信封、核心事件语义、surface 机制（`SurfaceEventType` 集合、`SurfaceOp` 变体）。拿不准就升：近似恒等的升级器几乎没有成本，漏升一次会让老读取器静默读坏。
+
+**读取规则按方向区分。**版本相等：正常读。比读取器新：拒绝，说明方向（"由更新的 harness 写入，请升级"），并给出原始日志文件的路径，用户仍能看到文本（`SessionFormatUnsupportedError`，与 `SessionPersistenceCorruptionError` 区分，因为数据没有损坏）。比读取器旧：查看时经 n→n+1 升级器链在内存中逐级转换；只有会话真正被继续时才把转换落盘（临时文件原子替换，原文件留备份）。写不出升级器的那一步留空，这会切断该步及更早所有版本的升级路径，它们降级为只能看原文。
+
+**逐事件的 `ignorable` 标记吸收词汇表增长，普通的新增事件永远不用升版本。**事件词汇表由挂载了哪些插件决定，单个版本整数描述不了它。读取器遇到不认识的事件类型时拒绝解读日志，除非该事件的信封带 `ignorable: true`。默认为必需：忘写标记的后果是把一个本可恢复的会话拒绝过头（体验问题），而默认可忽略会让同样的疏忽静默恢复出残缺会话（安全事故）。架构保证了这条规则成立：模型可见内容只经三种带 `surfaceOp` 标记的 surface 事件加 `request/header`、`request/context` 折叠进入重建，危险的未知事件恰好是那些不进 surface 但改变日志其余部分解读方式的事件（`session/end-seed` 是现存例子）。
+
+## 影响
+
+v0（0812 发布）交付的内容：分方向的拒绝并带原始日志路径；基于生成的已知词汇清单（`KNOWN_SESSION_EVENT_TYPES`，由 `gen-persistence-catalog` 从所有 `SessionEventMap` 声明合并生成，`verify-persistence-catalog` 保证新鲜）的未知事件守卫；`ignorable` 信封字段被种子校验、两个后端（SQLite 专用列，`SCHEMA_VERSION` 升到 15）和 BFF 线上 schema 接受。升级器链本身推迟到第一个真实的 v0→v1 变更出现、有真实对象可测时再建；写入侧目前不写 `ignorable`（还没有生产者需要它），`Session.append` 的这一表面随第一个使用者一起落地。在注册表面出现之前，仓库外插件的事件在第一方读取器下无法恢复会话，预发布立场接受这一点，而且拒绝是显式的而非静默的。未知类型守卫只在读取侧生效：`appendCore` 继续拒绝已淘汰的 legacy 形状，但不对新类型做词汇检查，因为写入时拒绝会让活跃会话的持久化中途停摆，代价大于下次加载时的显式拒绝。JSONL 后端还会在校验当前 header 形状、解码任何事件行之前，直接从原始 header 行拒绝外来版本，因此结构完全不同的未来格式仍会报告升级方向而不是"损坏"；SQLite 则先由自己的 `SCHEMA_VERSION` pragma 把关整个文件的结构。
+
+## 曾考虑的替代方案
+
+- **大小两级版本号**：能否转换这一位信息属于每一步的升级器，把它预先固化进编号形状会做出错误承诺。
+- **未知事件默认可忽略**：把忘写标记的后果从可见的过度拒绝反转成静默损坏。
+- **查看时自动迁移落盘**：打开即改写把读操作变成破坏性写操作，转换器的 bug 会在浏览时损坏日志，同目录的旧版本运行时也会因为新版本只是看了一眼就失去访问能力。
+- **插件运行时注册已知事件类型**：会让已知集依赖插件组合，同版本的精简组合会拒绝完整组合写出的日志。生成的全仓库清单保证同版本读取行为一致；仓库外插件的事件按构造就在清单之外，为它们提供注册表面推迟到真有这样的消费者时再做。
@@ -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/2026-08-11-trajectory-conversation-context-assembly.md
+2026-08-11-trajectory-conversation-context-assembly.md: 7d0aea2fc09f0f04bd5de923bee15c42a773489a
+2026-08-11-trajectory-conversation-context-assembly.zh.md: be3903bc62a8b1cc21bd9ddac450541ad4679ff0
@@ -0,0 +1,105 @@
+# Agent Note: Trajectory assembly from registered Conversation Contexts
+
+Status: implemented
+
+English | [中文](2026-08-11-trajectory-conversation-context-assembly.zh.md)
+
+## Problem
+
+Trajectory maintained an independent Session History source and folded the complete loaded Event window into Assistant, Tool, message, Request-header, and Compaction state. Chat already assembled the same Event families through registered Conversation Definitions. The two paths duplicated business correlation and pagination behavior, and a Trajectory structural update copied or rescanned work proportional to the raw Event count even when one business object changed.
+
+Reusing Chat's final Nodes would not solve the ownership problem. Trajectory needs request lifecycles, running Assistant state, prompt inheritance, Tool schemas, timing records, and a stage-oriented read model that Chat does not consume. Sharing final Node payloads would couple both views to the union of their requirements.
+
+The migration also had to preserve durable steering classification. A `user/message` does not say whether it opened a Turn or was claimed from the `next-step` inbox, and an older page can supply the missing inbox predecessor or Location after the message has already materialized.
+
+## Decision
+
+Trajectory registers target-owned Conversation Definitions and a `trajectory` View Builder against the shared [`ConversationNodeAssembler`](2026-08-09-client-conversation-node-assembly.md). Session owns one contiguous Event window and publishes both Chat and Trajectory snapshots through `Session.views`; it does not run a second Trajectory history source or business fold.
+
+Each Definition belongs to one target. Chat and Trajectory may recognize the same durable Event family, but they keep separate State and final Node payloads. They share only the Assembler's exact-ID matching, ordered Matches, Location facts, Reader dependencies, publication scheduling, and replace/prepend/append lifecycle.
+
+The existing [Trajectory inspection ledger](../feature/2026-07-27-trajectory-inspection-ledger.md) remains the view model. The Trajectory Builder converts materialized target Nodes into its established `eventNodes`, Requests, Tool schemas, running calls, and Location map; layout, table virtualization, selection, Overview, and inspector behavior do not become generic Conversation contracts.
+
+### Business Definitions
+
+| Business | Context identity | State assembly | Trajectory contribution |
+|---|---|---|---|
+| `next-step` inbox | splice Event seq | Apply the splice to the nearest preceding inbox Context | State only; no visible Node |
+| User, steering, or injected message | message Event seq | Read the preceding inbox State and classify the durable message | Input or context Node |
+| Assistant and ordinary Request | `turn:step` | Fold `step/start`, chunks, final message, retry, and `step/end` | Final Assistant, partial Assistant, and Request |
+| Root Tool call | root call ID | Fold root call/result and nested Code Dispatch events into one call tree | Final or running Tool tree |
+| Compaction | compaction ID | Fold start, summary, end, and replacement checkpoint | Compaction Request |
+| Request header | header Event seq | Read the preceding header and retain effective prompt plus the actual change | Prompt and Tool-schema source |
+| Session and Turn boundaries | boundary Event seq | Retain closure time and error facts | Interrupted Compaction or failed ordinary Request |
+
+Every correlating Event must expose the same business ID directly. Code Dispatch uses `rootCallId`, Compaction uses its compaction ID, and ordinary Tool and retry events retain their protocol identities even when a specific Definition correlates by `turn:step`. Legacy records that lack the required correlation ID are ignored by that Definition rather than merged into an `undefined` Context or crashing the Session.
+
+Assistant chunks update only their `turn:step` Context. Content-bearing chunks request animation-frame publication; usage and finish chunks update State without forcing their own frame. A final message, retry, or boundary publishes immediately. Completed Assistant State retains assembled blocks, timing, usage, and retry facts rather than copying the raw chunk ledger into the target snapshot.
+
+### Steering from predecessor Contexts
+
+Trajectory reconstructs steering from durable inbox history, using the same identity rule as the [Chat steering decision](../feature/2026-08-04-web-context-source-and-steer-marks.md) without sharing Chat's final Node.
+
+Each `agent/inbox/spliced` Event targeting `next-step` starts an invisible Context identified by its Event seq. Its `start()` reads the nearest earlier inbox Context, applies the splice, and stores the pending identities plus the cumulative set of claimed message IDs. A later user-origin `user/message` reads the nearest earlier inbox Context: a claimed ID produces a Steering Node, while every other user-origin message produces an ordinary User Node.
+
+A Reader miss while older history remains records a window-gap dependency. When prepend supplies the missing predecessor, the Assembler replays the affected inbox chain and message Contexts in forward Event order. Historical page direction therefore cannot permanently misclassify a message.
+
+The message Event's Location places steering in the owning Step. If the loaded history window lacks enough boundary Events to resolve that Location, layout uses the following Assistant step as the positional fallback. A running Request marker follows leading steering input in the same Step, so the marker denotes the model Request caused by that input rather than appearing before it.
+
+### Window paths and complexity
+
+Let `E` be the loaded raw Event count, `P` one newly prepended page, `D` the number of Trajectory Definitions, `C` the number of materialized Trajectory Context contributions, and `Mᵣ` the total Matches held by Contexts invalidated by a prepend. `D` is a small registered set; streaming chunks aggregate into one Assistant Context, so `C` is normally much smaller than `E`.
+
+| Path | Context work | Target snapshot work | Result |
+|---|---|---|---|
+| Initial tail or reconnect replace | Match the loaded window in `O(E × D)` and build State in forward Event order | Build and order `C` contributions | A full replace remains proportional to the loaded window |
+| Older-page prepend | Match only fresh Events and replay only Contexts whose Match, Location, or Reader answer changed, in `O(P × D + Mᵣ)` | Rebuild the stage snapshot from `C` contributions | Business folding does not restart over all `E` Events |
+| Live append | Match in `O(D)`, locate the keyed Context in `O(1)`, and update only that State | Replace a same-anchor contribution in `O(1)` before snapshot assembly | Business correlation is independent of loaded Event history |
+
+The Builder stores contributions by Context key and keeps a key-to-position index. A content update with the same anchor replaces one contribution in place; a new contribution or anchor change rebuilds and sorts contribution order. Snapshot assembly then walks `C` contributions, indexes Request headers and Tool schemas with Maps, and handles Compaction boundaries and Turn errors with linear cursors or indexes.
+
+Final Event and Request ordering keeps a publication's current upper bound at `O(C log C)`. The migration removes repeated reverse lookups and the old raw-history refold, but it does not claim end-to-end `O(1)` publication. Chat retains its existing keyed snapshot behavior and complexity; adding the Trajectory target does not make Chat scan Trajectory Contexts or Nodes.
+
+### Independent presentation hot paths
+
+The Context migration and the following presentation optimizations solve different costs. These reductions preserve the existing view model and are theoretical from call counts and asymptotic behavior; this decision does not claim benchmark measurements.
+
+| Hot path | Retained behavior | Expected reduction |
+|---|---|---|
+| Markdown summaries | Layout retains source Markdown; each stable Table record memoizes its displayed summary by content, while Detail parses only the selected record | A one-record append reparses the changed visible record instead of every Markdown record |
+| Search text | `TrajectorySearchIndex` linearly checks stable Record IDs and source signatures, but normalizes Markdown only for changed records and commits updates in three-second batches | Signature comparison remains `O(C)`; expensive normalization follows the changed-record count, and continuous frame updates collapse into one batch per interval |
+| Timeline tooltip | Timing text is computed after the delayed tooltip opens | A render with no open tooltip performs no per-span label formatting |
+| Following Assistant lookup | One reverse pass records the next Assistant for every input position | The former repeated forward lookup falls from worst-case `O(C²)` to `O(C)` |
+| Group duration | Fixed decimal grouping replaces `toLocaleString('en-US')` for the invariant English numeric shape | Complexity remains linear in Groups, but the Intl formatter leaves the repeated render path |
+
+Display memoization and search indexing stay separate. Search must include off-screen records and may lag live changes by the throttle interval; Table rendering must update the visible changed record immediately and must not inherit the index's commit cadence.
+
+## Alternatives considered
+
+**Keep the independent Session History fold and optimize it locally.** Rejected: caches could reduce selected hot paths, but Trajectory would still own a second Event window, pagination repair, request inspection fold, and business-correlation implementation beside Chat.
+
+**Reuse Chat Definitions and branch on a `target` argument in `buildViewNode()`.** Rejected: Trajectory needs different State and intermediate records, not only another React renderer. One Definition would carry both views' payloads and conditionals and would invalidate unrelated target data when either view changed.
+
+**Create a Trajectory-specific Assembler.** Rejected: exact-ID routing, update-before-start collection, prepend replay, Location repair, Reader dependencies, and publication cadence are not Trajectory-specific. A second engine would recreate the lifecycle duplication this change removes.
+
+**Add generic Surface, rewind, fanout, or settled lifecycle concepts.** Rejected: the current durable Event stream does not require a generic Surface branch, and Session or Turn boundaries are target business inputs rather than a reason to fan out one Event over every historical Context. Completion remains business State interpreted with Location closure.
+
+**Replace the Trajectory stages with generic Conversation Nodes.** Rejected: stages organize requests, timing, schemas, and table layout for one view. Making them engine contracts would constrain a future plain Session-log view and return view-specific composition to Client Runtime.
+
+**Share one Markdown cache between display and search.** Rejected: display is immediate and viewport-bound, while search covers the complete loaded record set and intentionally batches updates. A shared cache would couple correctness and scheduling across unrelated consumers.
+
+## Verification
+
+Runtime tests pin target registration, exact-ID append, update-before-start replay, prepend identity, Reader window-gap repair, Location replay, and isolation between Chat and Trajectory snapshots.
+
+Trajectory Definition and Builder tests pin Assistant streaming and interruption, nested Tool calls and parallel interruption, Compaction and prompt inheritance, Steering classification and Step placement, Request marker order, stable contribution replacement, and prepend expansion. Table, layout, Timeline, and search tests pin deferred Markdown work, throttled index updates, tooltip-time formatting, and stable search results across append and prepend.
+
+## Consequences
+
+Trajectory business assembly now scales with the changed page or keyed Context instead of restarting from the complete raw Event window. Target-owned Definitions can evolve independently from Chat while retaining one Session window and one set of lifecycle rules. Steering becomes a first-class Trajectory record at its actual Step position without adding steering-specific state to Session.
+
+The retained stage-oriented Builder still performs work proportional to materialized Trajectory contributions and may sort on publication. The search index still performs a light linear signature pass when its input layout changes. These costs are explicit target-view work, not hidden full Event refolding.
+
+Definition authors must provide stable protocol identities. Old Events without a required ID can disappear from the affected Trajectory business view, which is preferable to joining unrelated records or failing history load; producers that require faithful display must log the identity.
+
+The [Conversation assembly decision](2026-08-09-client-conversation-node-assembly.md) remains the authority for the generic Context, Reader, Location, and publication contracts. The [Trajectory ledger decision](../feature/2026-07-27-trajectory-inspection-ledger.md) remains the authority for table hierarchy, virtualization, inspector, and interaction behavior. This Note owns how Trajectory adapts those two decisions and why the adaptation does not share final Nodes with Chat.
@@ -0,0 +1,105 @@
+# Agent Note: Trajectory 基于注册式 Conversation Context 组装数据
+
+Status: implemented
+
+[English](2026-08-11-trajectory-conversation-context-assembly.md) | 中文
+
+## 问题
+
+Trajectory 曾维护独立的 Session History 数据源，并把完整的已加载 Event 窗口折叠为 Assistant、Tool、消息、Request header 和 Compaction 状态。Chat 已经通过注册式 Conversation Definition 组装相同的 Event 族。两条链路重复实现业务关联与分页行为；即使只改变一个业务对象，Trajectory 的结构更新仍会复制或重新扫描与原始 Event 数量成正比的数据。
+
+复用 Chat 的最终 Node 无法解决职责问题。Trajectory 需要请求生命周期、运行中 Assistant 状态、提示词继承、Tool schema、计时记录和 stage-oriented read model，而 Chat 不消费这些数据。共享最终 Node payload 会让两个视图都依赖双方需求的并集。
+
+本次迁移还必须保留持久 steering（中途引导）分类。`user/message` 本身不说明它是开启了一个 Turn，还是从 `next-step` inbox 被领取；更早页面还可能在消息已经物化后，才补齐缺失的 inbox 前驱或 Location。
+
+## 决策
+
+Trajectory 针对共享的 [`ConversationNodeAssembler`](2026-08-09-client-conversation-node-assembly.md) 注册 target 自有的 Conversation Definition 和 `trajectory` View Builder。Session 只维护一份连续 Event 窗口，并通过 `Session.views` 发布 Chat 与 Trajectory 快照；它不再运行第二套 Trajectory history source 或业务 fold。
+
+每个 Definition 只属于一个 target。Chat 与 Trajectory 可以识别同一持久 Event 族，但分别维护自己的 State 和最终 Node payload。它们只共享 Assembler 的精确 ID 匹配、有序 Match、Location 事实、Reader 依赖、发布调度，以及 replace/prepend/append 生命周期。
+
+既有的 [Trajectory 检查记录表](../feature/2026-07-27-trajectory-inspection-ledger.md)继续作为视图模型。Trajectory Builder 把已物化的 target Node 转换为原有的 `eventNodes`、Requests、Tool schema、运行中调用和 Location map；layout、表格虚拟化、选择、Overview 与检查器行为不会成为通用 Conversation 约定。
+
+### 业务 Definition
+
+| 业务 | Context 标识 | State 组装方式 | Trajectory contribution |
+|---|---|---|---|
+| `next-step` inbox | splice Event seq | 把 splice 应用到最近的前序 inbox Context | 只维护状态，不产生可见 Node |
+| 用户、steering 或注入消息 | message Event seq | 读取前序 inbox State，并对持久消息分类 | Input 或 context Node |
+| Assistant 与普通 Request | `turn:step` | 折叠 `step/start`、chunk、最终消息、retry 和 `step/end` | 最终 Assistant、partial Assistant 与 Request |
+| 根 Tool call | root call ID | 把根 call/result 与嵌套 Code Dispatch Event 折叠为一棵调用树 | 最终或运行中的 Tool tree |
+| Compaction | compaction ID | 折叠 start、summary、end 和 replacement checkpoint | Compaction Request |
+| Request header | header Event seq | 读取前一个 header，保留生效提示词及真实变化 | Prompt 与 Tool-schema 来源 |
+| Session 与 Turn 边界 | boundary Event seq | 保留关闭时间和错误事实 | 被中断的 Compaction 或失败的普通 Request |
+
+每个关联 Event 都必须直接提供相同的业务 ID。Code Dispatch 使用 `rootCallId`，Compaction 使用 compaction ID；即使某个 Definition 按 `turn:step` 关联，普通 Tool 与 retry Event 仍保留各自的协议标识。缺少必要关联 ID 的旧记录由该 Definition 忽略，不会合入 `undefined` Context，也不会导致 Session 崩溃。
+
+Assistant chunk 只更新对应的 `turn:step` Context。带内容的 chunk 请求 animation-frame 发布；usage 与 finish chunk 更新 State，但不单独强制刷新一帧。最终消息、retry 或边界立即发布。已完成 Assistant State 只保留组装后的 block、计时、usage 与 retry 事实，不会把原始 chunk ledger 复制进 target snapshot。
+
+### 通过前序 Context 恢复 steering
+
+Trajectory 从持久 inbox 历史恢复 steering，使用与 [Chat steering 决策](../feature/2026-08-04-web-context-source-and-steer-marks.md)相同的标识规则，但不共享 Chat 的最终 Node。
+
+每条目标为 `next-step` 的 `agent/inbox/spliced` Event 都会启动一个以 Event seq 标识的不可见 Context。它的 `start()` 读取最近的前序 inbox Context，应用 splice，并存储待处理标识以及累计的已领取 message ID 集合。后续用户来源的 `user/message` 读取最近的前序 inbox Context：已领取的 ID 生成 Steering Node，其余用户来源消息生成普通 User Node。
+
+仍有更早历史时，Reader miss 会记录 window-gap 依赖。prepend 补齐缺失的前驱后，Assembler 按 Event 正序重放受影响的 inbox chain 与 message Context。因此，历史分页方向不会永久错误分类消息。
+
+消息 Event 的 Location 会把 steering 放进所属 Step。如果已加载历史窗口缺少足够的边界 Event，无法解析该 Location，layout 就以后续 Assistant step 作为位置回退。同一个 Step 中，运行中 Request 标记排在前置 steering 输入之后，因此该标记表示由这条输入触发的模型 Request，而不会出现在输入前面。
+
+### 窗口链路与复杂度
+
+记 `E` 为已加载原始 Event 数，`P` 为一次新 prepend 的页面，`D` 为 Trajectory Definition 数，`C` 为已物化的 Trajectory Context contribution 数，`Mᵣ` 为一次 prepend 使其失效的 Context 所持有的 Match 总数。`D` 是较小的注册集合；流式 chunk 会聚合到同一个 Assistant Context，因此通常 `C` 明显小于 `E`。
+
+| 链路 | Context 工作量 | Target snapshot 工作量 | 结果 |
+|---|---|---|---|
+| 初始尾页或重连 replace | 以 `O(E × D)` 匹配已加载窗口，并按 Event 正序构造 State | 构造并排序 `C` 个 contribution | 完整 replace 仍与已加载窗口成正比 |
+| 更早页面 prepend | 只匹配新 Event，并只重放 Match、Location 或 Reader 答案发生变化的 Context，成本为 `O(P × D + Mᵣ)` | 从 `C` 个 contribution 重建 stage snapshot | 业务 fold 不会从头重跑全部 `E` 个 Event |
+| 实时 append | 以 `O(D)` 匹配，以 `O(1)` 找到 keyed Context，并只更新对应 State | snapshot 组装前，以 `O(1)` 替换 anchor 未变的 contribution | 业务关联成本与已加载 Event 历史无关 |
+
+Builder 按 Context key 保存 contribution，并维护 key-to-position index。anchor 相同的内容更新会原位替换一个 contribution；新增 contribution 或 anchor 变化才会重建并排序 contribution 顺序。随后，snapshot assembly 遍历 `C` 个 contribution，用 Map 索引 Request header 与 Tool schema，并以线性游标或索引处理 Compaction boundary 与 Turn error。
+
+最终 Event 和 Request 排序使单次发布的当前上界保持为 `O(C log C)`。本次迁移移除了重复反向查找和旧的原始历史 refold，但不声称端到端发布达到 `O(1)`。Chat 保持既有 keyed snapshot 行为与复杂度；增加 Trajectory target 不会让 Chat 扫描 Trajectory Context 或 Node。
+
+### 独立的表现层热点优化
+
+Context 迁移与下列表现层优化解决的是不同成本。这些优化保留既有视图模型；收益来自调用次数和渐进复杂度推算，本决策不声称存在 benchmark 实测结果。
+
+| 热点 | 保留的行为 | 预期减少的工作 |
+|---|---|---|
+| Markdown 摘要 | Layout 只保留源 Markdown；每个稳定 Table record 按内容 memo 展示摘要，Detail 只解析当前选中记录 | 单条 record append 只重解析发生变化的可见记录，而非全部 Markdown record |
+| 搜索文本 | `TrajectorySearchIndex` 仍线性核对稳定 Record ID 与来源签名，但只为变化的 record 标准化 Markdown，并以三秒批次提交更新 | 签名比较仍为 `O(C)`；昂贵标准化只随变化 record 数量增长，持续 frame update 每个时间窗合并成一个批次 |
+| Timeline tooltip | 延迟 Tooltip 打开后才计算计时文案 | 没有打开 Tooltip 的 render 不执行逐 span label 格式化 |
+| 后继 Assistant 查找 | 一次反向遍历为每个输入位置记录后续 Assistant | 原先重复向前查找的最坏复杂度从 `O(C²)` 降为 `O(C)` |
+| Group duration | 以固定十进制分组替代固定英文数字形态下的 `toLocaleString('en-US')` | 复杂度仍与 Group 数线性相关，但重复 render 路径不再调用 Intl formatter |
+
+展示 memo 与搜索索引彼此独立。搜索必须覆盖屏幕外 record，并允许实时变化延迟一个 throttle 周期；Table 必须立即更新发生变化的可见 record，不能继承索引的提交节奏。
+
+## 考虑过的替代方案
+
+**保留独立 Session History fold，只做局部优化。** 不予采纳：缓存可以降低部分热点，但 Trajectory 仍会在 Chat 之外拥有第二套 Event 窗口、分页修复、request inspection fold 与业务关联实现。
+
+**复用 Chat Definition，并在 `buildViewNode()` 中按 `target` 分支。** 不予采纳：Trajectory 需要不同的 State 与中间 record，不只是另一套 React renderer。单一 Definition 会携带两个视图的 payload 与条件，并在任一视图变化时让无关 target 数据失效。
+
+**创建 Trajectory 专属 Assembler。** 不予采纳：精确 ID 路由、先 update 后 start 的收集、prepend replay、Location 修复、Reader 依赖与发布节奏都不是 Trajectory 特有行为。第二套引擎会重新制造本次改造要消除的生命周期重复。
+
+**增加通用 Surface、rewind、fanout 或 settled 生命周期。** 不予采纳：当前持久 Event stream 不需要通用 Surface branch；Session 或 Turn boundary 是 target 业务输入，不构成把一个 Event fanout 到全部历史 Context 的理由。完成条件仍由业务 State 结合 Location closure 判断。
+
+**用通用 Conversation Node 替换 Trajectory stage。** 不予采纳：stage 为单一视图组织 Request、计时、schema 和表格 layout。把它变成引擎约定会限制未来的朴素 Session-log 视图，并把视图专属组合重新放回 Client Runtime。
+
+**在展示与搜索之间共享一套 Markdown cache。** 不予采纳：展示要求立即更新且受 viewport 约束，搜索则覆盖全部已加载 record，并有意批量提交更新。共享 cache 会把两个无关消费方的正确性与调度节奏耦合起来。
+
+## 验证
+
+Runtime 测试固定 target 注册、精确 ID append、先 update 后 start 的 replay、prepend identity、Reader window-gap 修复、Location replay，以及 Chat 与 Trajectory snapshot 隔离。
+
+Trajectory Definition 与 Builder 测试固定 Assistant streaming 与 interruption、嵌套 Tool call 和并行 interruption、Compaction 与 prompt 继承、Steering 分类和 Step 位置、Request 标记顺序、稳定 contribution 替换与 prepend 扩展。Table、layout、Timeline 与搜索测试固定延迟 Markdown 工作、节流索引更新、Tooltip 展示时格式化，以及 append/prepend 期间稳定的搜索结果。
+
+## 后果
+
+Trajectory 业务组装的成本随变化页面或 keyed Context 增长，不再从完整原始 Event 窗口重新开始。target 自有 Definition 可以独立于 Chat 演进，同时继续共享一份 Session 窗口和一套生命周期规则。steering 会在实际所属 Step 位置成为一等 Trajectory record，不需要向 Session 增加 steering 专属状态。
+
+保留的 stage-oriented Builder 仍会执行与已物化 Trajectory contribution 数量成正比的工作，并可能在发布时排序。输入 layout 变化时，搜索索引仍会执行一次轻量线性签名检查。这些成本是显式的 target view 工作，不是隐藏的完整 Event refold。
+
+Definition 作者必须提供稳定的协议标识。缺少必要 ID 的旧 Event 可能不会出现在受影响的 Trajectory 业务视图中；与合并无关记录或让历史加载失败相比，这是更安全的退化方式。要求完整展示的生产方必须记录该标识。
+
+[Conversation assembly 决策](2026-08-09-client-conversation-node-assembly.md)继续作为通用 Context、Reader、Location 与发布约定的真源。[Trajectory ledger 决策](../feature/2026-07-27-trajectory-inspection-ledger.md)继续负责表格层级、虚拟化、检查器和交互行为。本 Note 负责说明 Trajectory 如何适配这两项决策，以及为何该适配不与 Chat 共享最终 Node。
@@ -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/2026-08-03-cli-signal-shutdown-escalation.md
-2026-08-03-cli-signal-shutdown-escalation.md: 55917400fac2728d13dc2cdd799a7e234b6ed661
-2026-08-03-cli-signal-shutdown-escalation.zh.md: c7897a8d77e8c2ebad43cec4e12170b04c837350
+2026-08-03-cli-signal-shutdown-escalation.md: 173d06482cd8a1fcbb985763cc313e3f9b170bc6
+2026-08-03-cli-signal-shutdown-escalation.zh.md: efa52524199906cf636cb2b55cb4249857dc1348
@@ -6,7 +6,7 @@ English | [中文](2026-08-03-cli-signal-shutdown-escalation.zh.md)

 ## Problem

-The default telemetry mount added SIGINT/SIGTERM handlers to `dsh web` and the headless command (now `dsh run`) so process exit could drain the Cordis tree instead of dropping queued telemetry. Each handler used a one-way boolean latch and exited only after `ctx.fiber.dispose()` settled. Headless normal completion also awaited that disposal without a bound.
+The default telemetry mount added SIGINT/SIGTERM handlers to `dsh web` and the headless command (now `dsh --profile headless`) so process exit could drain the Cordis tree instead of dropping queued telemetry. Each handler used a one-way boolean latch and exited only after `ctx.fiber.dispose()` settled. Headless normal completion also awaited that disposal without a bound.

 A user then reproduced the headless command hanging immediately after the observation URL and ignoring repeated `Ctrl+C`; `DSH_TELEMETRY_DISABLED=1` removed the hang, while a standalone Node handler in the same Linux sandbox received SIGINT. This isolated the pending disposer to telemetry rather than terminal signal forwarding. OTel's `BatchLogRecordProcessor.shutdown()` awaits `exporter.forceFlush()` before the `exportTimeoutMillis`-bounded completion promise, and the OTLP exporter's `forceFlush()` waits directly on its in-flight HTTP Promise. A proxy/sandbox connection that never obtains a socket can therefore leave provider shutdown pending despite both configured SDK timeouts.

@@ -6,7 +6,7 @@ Status: implemented

 ## 问题

-默认挂载遥测后，`dsh web` 与 headless 命令（现为 `dsh run`）新增了 SIGINT/SIGTERM 处理器，使进程退出时可以排空 Cordis 插件树，而不是丢弃排队中的遥测数据。每个处理器都使用单向布尔闩锁（latch），并且只有在 `ctx.fiber.dispose()` 结算后才退出。headless 正常完成时同样会无界等待整棵树执行 dispose（资源释放）。
+默认挂载遥测后，`dsh web` 与 headless 命令（现为 `dsh --profile headless`）新增了 SIGINT/SIGTERM 处理器，使进程退出时可以排空 Cordis 插件树，而不是丢弃排队中的遥测数据。每个处理器都使用单向布尔闩锁（latch），并且只有在 `ctx.fiber.dispose()` 结算后才退出。headless 正常完成时同样会无界等待整棵树执行 dispose（资源释放）。

 随后有用户复现，headless 命令在打印观察 URL 后立即卡死，重复按 `Ctrl+C` 也没有反应；设置 `DSH_TELEMETRY_DISABLED=1` 后不再卡死，而同一 Linux 沙箱中的独立 Node 信号处理器能够收到 SIGINT。这将待结算的 disposer 定位到遥测，而非终端信号转发。OTel 的 `BatchLogRecordProcessor.shutdown()` 会先等待 `exporter.forceFlush()`，再进入受 `exportTimeoutMillis` 限制的完成 promise；OTLP 导出器的 `forceFlush()` 则直接等待正在进行的 HTTP Promise。因此，代理／沙箱连接始终无法取得 socket 时，即使已经配置两项 SDK 超时，也会让提供方关闭一直待结算。

@@ -1,6 +0,0 @@
-# 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/2026-08-08-npm-backed-git-repository-plugin-preparation.md
-2026-08-08-npm-backed-git-repository-plugin-preparation.md: 958b932f82f4da3cf63aa911260411855e514409
-2026-08-08-npm-backed-git-repository-plugin-preparation.zh.md: d2256e0eae303c371371b9b5ba1967105aa61834
@@ -1,48 +0,0 @@
-# Agent Note: npm-backed preparation makes GitHub repository Plugins self-contained
-
-Status: implemented
-
-English | [中文](2026-08-08-npm-backed-git-repository-plugin-preparation.zh.md)
-
-## Problem
-
-The repository Plugin authoring contract requires `scripts.prepack` to invoke `dsh-plugin-prepare`. Supplying that executable from the running DSH installation made a source package appear valid even when its own manifest could not obtain the helper. It therefore did not prove the behavior users need after `@deepseek-ai/dsh-repository-plugin` is published: an ordinary Git-hosted npm package must be installable and preparable from only its declared dependencies.
-
-A selectable `.dsh-plugin` inside a pnpm workspace has a second isolation requirement. pnpm prepares a Git-hosted package by running the repository's preferred package manager before packing the selected subdirectory. A nested `pnpm install` can join the containing workspace; when the root lockfile does not list `.dsh-plugin` as an importer, pnpm can report success without installing dependencies declared only by that package. Its TypeScript build or prepare command then fails, or a pre-generated artifact hides the missing dependency.
-
-The checked-in headless fixture mounts an already prepared wrapper. It proves runtime composition, not GitHub acquisition, npm resolution, or package-owned preparation.
-
-## Decision
-
-The `.dsh-plugin` package declares `@deepseek-ai/dsh-repository-plugin` as an ordinary development dependency and invokes its published `dsh-plugin-prepare` executable from `scripts.prepack`. The package may declare any other build and runtime dependencies and run arbitrary compilation before the helper. The repository Plugin package marks its Cordis and DSH peers optional so a helper-only development install resolves only the helper's actual `zod` runtime dependency; an application composition still supplies the peers used by the package's Cordis entry.
-
-DSH does not materialize or prepend a prepare executable. `RepositoryCache` supplies only a transaction-owned `pnpm` wrapper: the outer install runs the pinned pnpm entry directly, while pnpm's hard-coded Git-package `pnpm install` reinvokes the same entry with `--ignore-workspace`. The selected package therefore owns dependency resolution even beneath another pnpm lockfile, and normal package-manager lifecycle `PATH` construction exposes `node_modules/.bin/dsh-plugin-prepare`. The temporary pnpm wrapper disappears after the child settles. The repository remains trusted package-manager input: all dependency and lifecycle code executes under the existing trust contract.
-
-The Node 24 consumer lane passes an exact source derived from the pull request head repository and SHA. It uses the existing private DeepSeek Harness repository rather than creating another repository per run. A job-scoped Git configuration gives the read-only job token access to that exact private source and rewrites pnpm's SSH fallback to authenticated HTTPS.
-
-The built-entry acceptance also creates an in-process npm registry. It stages the current built `@deepseek-ai/dsh-repository-plugin` as a publication artifact by removing `private`, replacing workspace protocols with the release version, and packing the declared files. The registry serves the resulting packument and tarball, while a job-local npm config directs only the `@deepseek-ai` scope to it. The real built `dsh run` child then fetches the exact Git source; that package resolves the helper through npm, type-checks and bundles a TypeScript Cordis entry and MCP server, prepares the adjacent skill, and loads all three contributions. A deliberately failing host `PATH` command proves the lifecycle selected the dependency-local executable. The acceptance also requires registry resolution and inspects the immutable prepared cache, so restoring a host-injected helper cannot satisfy it.
-
-## Alternatives considered
-
-**Inject `dsh-plugin-prepare` from the running DSH installation.** Rejected because it lets an incomplete repository manifest pass and tests a host-only path that npm consumers cannot reproduce.
-
-**Publish the source fixture itself to npm.** Rejected because the product contract is specifically that the DSH Plugin remains Git-hosted; only the reusable preparation helper is an npm dependency.
-
-**Create a new private GitHub repository in every CI run.** Rejected because the pull request repository at its exact head SHA is already a real authenticated private Git remote. Per-run repository mutation would add credentials, cleanup, and eventual-consistency failure modes without changing the acquisition path.
-
-**Prepare after `RepositoryCache` installs the selected package.** Rejected because pnpm's packed subdirectory no longer contains sibling source assets referenced by paths such as `../skills`; preparation must happen before packlist.
-
-**Clone GitHub repositories in DSH and bypass pnpm's Git fetcher.** Rejected because it would duplicate ref resolution, subdirectory selection, dependency installation, packlist behavior, and cache integrity already owned by the pinned package manager.
-
-## Consequences
-
- A repository author can commit a `.dsh-plugin` package, TypeScript source, skills, and MCP definitions to GitHub without publishing that Plugin package to npm. The package must declare the published preparation dependency.
- Private GitHub sources use the host's standard Git authentication. CI proves that path with a temporary read-only configuration rather than persistent runner credentials.
- `prepack`, not `prepare`, is part of the authoring format. It may contain arbitrary package-owned build steps but must invoke the dependency-provided helper; missing dependency or lifecycle metadata fails before a cache generation is usable.
- A selected package in a pnpm repository installs from its own manifest rather than an enclosing workspace. It cannot rely on workspace-only hoisting; ordinary registry and relative `file:` dependencies remain package-owned inputs.
- Exact source strings identify immutable cache generations; a changed ref or source configuration selects another generation.
- Package dependencies, compilation, preparation, and the trusted `dsh.entry` contribution remain owned by the repository package and the [trusted-code decision](../architecture/2026-08-08-trusted-repository-package-code.md).
-
-## Testing
-
-`packages/boot/app-boot/tests/repository-cache.spec.ts` runs a package excluded from its source repository's root pnpm lockfile through a local Git subpath and requires relative `file:` dependencies to provide both its build command and `dsh-plugin-prepare`; it also proves that visible environment survives while credential-shaped variables are scrubbed. `packages/self-modification/repository-plugin/tests/repository-plugin.spec.ts` pins helper-bearing `prepack` metadata and preparation output. `examples/headless-agent/tests/keyless-smoke.e2e.ts` keeps the checked-in prepared fixture on that source contract. `apps/cli/tests/github-repository-plugin.built.e2e.ts` is the product acceptance: simulated published helper package, job-local npm registry, fresh DSH home, exact authenticated private GitHub source, actual built `dsh run`, package-owned TypeScript build, real MCP execution, code-entry transformation, mock LLM request observation, and prepared cache inspection.
@@ -1,48 +0,0 @@
-# Agent Note: 基于 NPM 的准备机制使 GitHub repository 插件自包含
-
-状态：已实现
-
-[English](2026-08-08-npm-backed-git-repository-plugin-preparation.md) | 中文
-
-## 问题
-
-repository 插件创作约定要求 `scripts.prepack` 调用 `dsh-plugin-prepare`。如果由正在运行的 DSH 安装提供该可执行文件，即使源包自身的 manifest（元数据清单）无法取得辅助程序，它也会显得有效。因此，这并未证明 `@deepseek-ai/dsh-repository-plugin` 发布后用户所需的行为：普通 Git 托管 NPM 包必须只依靠自身声明的依赖即可安装和准备。
-
-pnpm workspace 内可选择的 `.dsh-plugin` 还有另一项隔离要求。pnpm 会在打包所选子目录前运行仓库首选的包管理器，以准备 Git 托管包。嵌套的 `pnpm install` 可能加入外层 workspace；当根 lockfile 未把 `.dsh-plugin` 列为 importer 时，pnpm 可能报告成功，却未安装仅由该包声明的依赖。随后，其 TypeScript 构建或准备命令会失败；也可能因为存在预生成产物，依赖缺失被掩盖。
-
-签入仓库的 headless fixture（测试前置数据）挂载的是已准备好的包装层。它证明运行时组合，而不证明 GitHub 获取、NPM 解析或包自有准备。
-
-## 决策
-
-`.dsh-plugin` 包将已发布的 `@deepseek-ai/dsh-repository-plugin` 声明为普通开发依赖，并在 `scripts.prepack` 中调用其已发布的 `dsh-plugin-prepare` 可执行文件。该包可以声明其他任意构建依赖与运行时依赖，并在辅助程序前执行任意编译。repository 插件包把 Cordis 与 DSH 对等依赖（peer dependency）标为可选，因此仅为使用辅助程序而进行的开发安装只会解析辅助程序实际依赖的 `zod` 运行时依赖；应用组合仍会提供该包 Cordis 入口所使用的对等依赖。
-
-DSH 不会生成准备阶段可执行文件，也不会将其前置到 `PATH`。`RepositoryCache` 只提供一个由事务持有的 `pnpm` 包装脚本：外层安装直接运行锁定的 pnpm 入口，而 pnpm 为 Git 包硬编码的 `pnpm install` 会以 `--ignore-workspace` 重新调用同一入口。因此，即使位于另一个 pnpm lockfile 之下，所选包仍自行负责依赖解析，正常的包管理器生命周期 `PATH` 构造会暴露 `node_modules/.bin/dsh-plugin-prepare`。临时 pnpm 包装脚本会在子进程结算后消失。repository 仍是受信任的包管理器输入：所有依赖与生命周期代码都按既有信任约定执行。
-
-Node 24 消费方 CI 任务会传入从 PR（Pull Request）head 仓库与 SHA 派生的精确源。它复用现有私有 DeepSeek Harness 仓库，而不会为每次运行新建仓库。作业作用域的 Git 配置允许只读作业 token 访问该精确私有源，并把 pnpm 的 SSH 回退改写为已认证 HTTPS。
-
-构建入口验收还会创建一个进程内 NPM 注册表。它通过移除 `private`、将 workspace protocol 替换为发布版本并打包声明的文件，把当前已构建的 `@deepseek-ai/dsh-repository-plugin` 暂存为发布产物。注册表会提供由此生成的 `packument` 与 tarball，作业本地 NPM 配置则只把 `@deepseek-ai` scope 指向它。实际构建的 `dsh run` 子进程随后获取精确 Git 源；该包通过 NPM 解析辅助程序，对 TypeScript Cordis 入口和 MCP server 进行类型检查与打包，准备相邻的 skill（技能），并加载全部三类贡献。一个刻意设为失败的宿主 `PATH` 命令可以证明，该生命周期选中的是依赖内的可执行文件。验收还要求经过注册表解析并检查不可变的已准备缓存，因此恢复宿主注入的辅助程序也无法通过。
-
-## 考虑过的替代方案
-
-**从正在运行的 DSH 安装注入 `dsh-plugin-prepare`。** 拒绝，因为这会让 manifest 不完整的 repository 包通过，并测试 NPM 消费方无法复现的纯宿主路径。
-
-**把源 fixture 本身发布到 NPM。** 拒绝，因为产品约定明确要求 DSH 插件仍托管在 Git；只有可复用的准备辅助程序是 NPM 依赖。
-
-**在每次 CI 运行中创建新的私有 GitHub 仓库。** 拒绝，因为 PR 仓库的精确 head SHA 已是经过认证的真实私有 Git remote。每次运行的仓库变更会增加凭据、清理和最终一致性失败模式，却不改变获取路径。
-
-**在 `RepositoryCache` 安装所选包后再准备。** 拒绝，因为 pnpm 打包后的子目录不再包含 `../skills` 等路径所引用的同仓库相邻资源；准备必须在生成 packlist 前完成。
-
-**在 DSH 中克隆 GitHub 仓库并绕过 pnpm 的 Git 获取器。** 拒绝，因为这会重复实现已由锁定包管理器负责的 ref 解析、子目录选择、依赖安装、packlist 行为和缓存完整性。
-
-## 后果
-
- 仓库作者可以把 `.dsh-plugin` 包、TypeScript 源码、skill 与 MCP 定义提交到 GitHub，而无需把该插件包发布到 NPM。该包必须声明已发布的准备依赖。
- 私有 GitHub 源使用宿主的标准 Git 认证。CI 使用临时的只读配置而非运行器上的持久凭据来验证该路径。
- 创作格式使用 `prepack` 而不是 `prepare`。其中可以包含任意包自有构建步骤，但必须调用依赖提供的辅助程序；依赖或生命周期元数据缺失时，会在缓存 generation 可用前失败。
- pnpm 仓库中的所选包按自身 manifest 安装，而不继承外层 workspace。它不能依赖仅由 workspace 提升而可见的包；普通注册表依赖和相对 `file:` 依赖仍是包自有输入。
- 精确源字符串标识不可变缓存 generation；改变 ref 或源配置会选择另一个 generation。
- 包依赖、编译、准备和受信任的 `dsh.entry` 贡献仍由 repository 包和[受信任代码决策](../architecture/2026-08-08-trusted-repository-package-code.md)负责。
-
-## 测试
-
-`packages/boot/app-boot/tests/repository-cache.spec.ts` 会通过本地 Git 子路径运行一个未列入源仓库根 pnpm lockfile 的包，并要求相对 `file:` 依赖同时提供构建命令与 `dsh-plugin-prepare`；该测试还证明可见环境变量得以保留，而名称符合凭据模式的变量会被清除。`packages/self-modification/repository-plugin/tests/repository-plugin.spec.ts` 锁定包含辅助命令的 `prepack` 元数据与准备输出。`examples/headless-agent/tests/keyless-smoke.e2e.ts` 使签入仓库的已准备 fixture 继续符合该源格式约定。`apps/cli/tests/github-repository-plugin.built.e2e.ts` 是产品验收测试：模拟发布的辅助程序包、作业本地 NPM 注册表、全新 DSH 主目录、精确且经过认证的私有 GitHub 源、实际构建的 `dsh run`、包自有 TypeScript 构建、真实 MCP 执行、代码入口转换、mock LLM（大语言模型）请求观测，以及已准备缓存检查。
@@ -1,6 +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/2026-07-30-static-repository-plugin-format.md
-2026-07-30-static-repository-plugin-format.md: c66ee111eb0cac9e0d6c54581855ffc18efc8611
-2026-07-30-static-repository-plugin-format.zh.md: c85aaf44d96098eb1ccb45456cce1a15e7408fc5
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-08-10-minimal-preset-owns-rl-composition.md
+2026-08-10-minimal-preset-owns-rl-composition.md: 002cad0827e969b322997821dc978db85e2955f3
+2026-08-10-minimal-preset-owns-rl-composition.zh.md: e957b57395c68b336695bdae07ea15a54ca1ea4e
@@ -0,0 +1,39 @@
+# Agent Note: The minimal preset owns the complete RL agent composition
+
+Status: implemented
+
+English | [中文](2026-08-10-minimal-preset-owns-rl-composition.zh.md)
+
+## Problem
+
+The Web surface offered two owners for the Claude SWE-compatible RL agent: a process-wide `core-web.cordis.yml` patch and the per-session `minimal` preset. Once [agent presets](../architecture/2026-08-03-per-session-agent-presets.md) became the agent-composition boundary, the preset's scoped `deployment:persona` shadowed the overlay's corrected global persona with stale coding-agent text. The overlay test mounted no preset, while the preset test booted without the overlay, so neither exercised the composition users selected.
+
+The split also hid other drift. The preset mounted one-shot Bash rather than the [persistent Bash](../feature/2026-07-29-persistent-bash-str-replace-editor.md) used by the RL harness and omitted the RL compaction policy. Keeping both owners makes every future prompt, tool, and policy change a cross-product.
+
+## Decision
+
+The shipped Web `minimal` preset is the sole Web owner of the RL agent composition. It declares an entry-local PTY registry and local backend, persistent `bash` with the RL environment description and 300-second timeout, `str_replace_editor`, and an entry-local compaction backend. Tool presentation remains a deployment choice. The compaction policy keeps the RL threshold, absolute retention, generation cap, and retry count; model capacity comes from routed adapter metadata because `contextWindow` is no longer a compact-basic config field. The editor accepts no `requireAbsolutePath` setting because absolute paths are its unconditional contract.
+
+The preset persona is exactly `You are a helpful software engineer assistant.` and sets `complete: true`. A complete `PromptSection` participates in ordinary assembly so tools, contexts, variables, and cooperative listeners still resolve; after the `system-prompt/assemble` waterfall, the prompt registry restores a detached copy of that section as the sole system-prompt section. Multiple effective complete sections reject assembly. This final registry constraint prevents harness identity, Web orientation, tool guidance, or an assembly listener from appending prompt text.
+
+The process-wide `core-web.cordis.yml` patch is absent. Browser UI, workspace attachment, persistence, filesystem, subprocess, sandbox, permission, model routing, and other cross-session services remain host-owned. Selecting `minimal` changes one agent's model-facing composition without changing other sessions in the Web process.
+
+## Verification
+
+System-prompt and persona package tests prove final complete-section enforcement, including waterfall mutation and duplicate rejection. The shipped-preset composition test asserts the exact prompt, Bash description, absolute editor schema, and two-tool catalog under the default native presentation. The keyless Web replay sends a real request through a `minimal` agent while global identity, Web surface text, and a test section are registered, then executes two persistent Bash calls to prove environment and cwd state survive and executes the editor through an absolute path.
+
+The standalone [`minimal.cordis.yml`](../../../../examples/jsonrpc-agent/minimal.cordis.yml) mirrors the same prompt, tools, timeouts, and compaction policy for the bundled JSON-RPC runtime. Its keyless SDK replay asserts the assembled system prompt and two-tool catalog, executes persistent Bash across calls, and exercises the editor; the Python SDK tutorial provides the runnable entry point.
+
+## Alternatives considered
+
+**Keep `core-web.cordis.yml` as a compatibility patch.** Rejected because a process patch and a session preset are two independent owners for one agent contract; precedence makes either one capable of silently undoing the other.
+
+**Disable every known prompt contributor in the preset.** Rejected because host rows are process-wide and new contributors would reopen the prompt. A final complete-section constraint expresses the negative guarantee at the registry that assembles the prompt.
+
+**Filter sections only with a prepended waterfall listener.** Rejected because another prepended wrapper can run outside it and append after the filter. Enforcement after the complete waterfall has stable final authority.
+
+**Mount PTY services on the Web host.** Rejected because only the minimal agent consumes them. An entry-local `pty` realm gives the services the same lifetime and scope as their sole consumer without publishing a process-global service from a preset.
+
+## Consequences
+
+The RL prompt is fixed rather than environment-overridable. The Web preset and standalone JSON-RPC example state the same contract for their respective launch surfaces. The model sees only persistent `bash` and `str_replace_editor`; shell state is per agent and disappears with that agent. The preset pays for its own PTY and compaction service instances, while other presets pay nothing for them. The local persistent-shell backend requires the supported POSIX terminal substrate, so this preset is not a Windows agent surface.
@@ -0,0 +1,39 @@
+# Agent Note: minimal preset 拥有完整的 RL agent 组合
+
+Status: implemented
+
+[English](2026-08-10-minimal-preset-owns-rl-composition.md) | 中文
+
+## 问题
+
+Web surface 同时由两个位置定义与 Claude SWE 兼容的 RL agent（智能体）：进程级 `core-web.cordis.yml` patch，以及逐会话的 `minimal` preset。[agent preset](../architecture/2026-08-03-per-session-agent-presets.md) 成为 agent 组合边界后，preset 中带作用域的 `deployment:persona` 会用陈旧的 coding-agent 文本遮蔽 overlay 修正过的全局 persona。overlay 测试没有挂载 preset，而 preset 测试启动时没有 overlay，因此两者都没有覆盖用户实际选择的组合。
+
+这种拆分还掩盖了其他偏差。preset 挂载了一次性 Bash，而不是 RL harness 使用的[持久 Bash](../feature/2026-07-29-persistent-bash-str-replace-editor.md)，并且遗漏了 RL 压缩（compaction）策略。保留两个所有者，会使今后每次修改提示词、工具或策略时都必须验证二者的交叉组合。
+
+## 决策
+
+随附的 Web `minimal` preset 是 RL agent 组合在 Web 中的唯一所有者。它声明 entry 本地的 PTY 注册表与本地后端、带 RL 环境描述且超时为 300 秒的持久 `bash`、`str_replace_editor`，以及 entry 本地的压缩后端。工具呈现仍由部署选择。压缩策略保留 RL 的阈值、绝对保留量、生成上限和重试次数；模型容量来自经路由选定的适配器元数据，因为 `contextWindow` 已不再是 compact-basic 的配置字段。编辑器不接受 `requireAbsolutePath` 设置，因为要求绝对路径是它的无条件约定。
+
+preset persona 恰好是 `You are a helpful software engineer assistant.`，并设置 `complete: true`。complete `PromptSection` 参与常规组装，因此工具、上下文、变量和协作式监听器仍会解析；`system-prompt/assemble` waterfall（瀑布式事件）结束后，提示词注册表会将该段落的独立副本恢复为唯一的系统提示词段落。存在多个有效 complete 段时，组装会被拒绝。这项最终注册表约束可防止 harness 身份、Web 定位、工具引导或组装监听器追加提示词文本。
+
+进程级 `core-web.cordis.yml` patch 不再存在。浏览器 UI、workspace 附加、持久化、文件系统、子进程、沙箱、权限、模型路由及其他跨会话服务仍由宿主持有。选择 `minimal` 只会改变一个 agent 面向模型的组合，不会改变 Web 进程中的其他会话。
+
+## 验证
+
+系统提示词与 persona 包测试证明了 complete 段的最终约束，包括 waterfall 修改与重复项拒绝。交付 preset 组合测试在默认原生呈现下断言精确的提示词、Bash 描述、要求绝对路径的编辑器 schema 和双工具目录。无密钥 Web 回放通过 `minimal` agent 发送一个真实请求，同时注册全局身份、Web surface 文本和一个测试段落；随后执行两次持久 Bash 调用，证明环境与 cwd 状态能够保留，并通过绝对路径执行编辑器。
+
+独立的 [`minimal.cordis.yml`](../../../../examples/jsonrpc-agent/minimal.cordis.yml) 为内置 JSON-RPC 运行时复现相同的提示词、工具、超时和压缩策略。其无密钥 SDK 回放会断言组装后的系统提示词与双工具目录，跨调用执行持久 Bash，并使用编辑器；Python SDK 教程提供可运行的入口。
+
+## 考虑过的替代方案
+
+**将 `core-web.cordis.yml` 保留为兼容 patch。** 被拒绝，因为进程 patch 与会话 preset 是同一 agent 约定的两个独立所有者；优先级会使任意一方都能静默撤销另一方的配置。
+
+**在 preset 中禁用每个已知的提示词贡献方。** 被拒绝，因为宿主行属于整个进程，新的贡献方也会重新开放提示词。由组装提示词的注册表实施最终 complete 段约束，才能表达这项否定保证。
+
+**仅使用前置 waterfall 监听器筛选段落。** 被拒绝，因为另一个前置包装层可以在该监听器外执行，并在筛选后追加内容。在整个 waterfall 结束后实施约束，才能稳定拥有最终决定权。
+
+**在 Web 宿主上挂载 PTY 服务。** 被拒绝，因为只有 minimal agent 消费这些服务。entry 本地的 `pty` realm 与唯一消费方具有相同的生命周期和作用域，无需由 preset 发布进程级全局服务。
+
+## 后果
+
+RL 提示词固定不变，不能通过环境覆盖。Web preset 与独立 JSON-RPC 示例分别在各自的启动界面声明相同的约定。模型只看到持久 `bash` 与 `str_replace_editor`；shell 状态按 agent 隔离，并随该 agent 一并消失。preset 为自身的 PTY 与压缩服务实例承担开销，其他 preset 无需承担。持久 shell 的本地后端需要受支持的 POSIX 终端基础环境，因此该 preset 不适用于 Windows agent surface。
@@ -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/2026-08-10-subagent-empty-terminal-message-output.md
+2026-08-10-subagent-empty-terminal-message-output.md: 693013f6810005ce02b08bd82f1f6a18511c40fb
+2026-08-10-subagent-empty-terminal-message-output.zh.md: 64d61af21f838ef3f515db8af116cbdd74e96179
@@ -0,0 +1,31 @@
+# Agent Note: One selection rule keeps subagent output past an empty terminal message
+
+Status: implemented
+
+English | [中文](2026-08-10-subagent-empty-terminal-message-output.zh.md)
+
+## Problem
+
+The agent loop appends an empty-content `assistant/message` when a `max-tokens` step assembled only tool-call blocks because `BlockAssembler.blocks()` drops truncated tool calls; the message records usage only. Three consumers selected the child's output independently and treated that usage record as output. The in-process driver's `readResult` and the continuable Activation's `subagent/end` capture selected the last `assistant/message` without filtering, while the SDK backend's observer let any `assistant/message` take precedence over accumulated text. In a multi-step turn cut off at max-tokens, the final empty message caused the real partial answer to be omitted from `SubagentResult.output`, the tool result, telemetry, and `subagent/end.lastAssistantMessage`. The in-process driver also lacked a streamed-text fallback, so a cancelled child whose only text existed in `assistant/chunk` events reported `[]`.
+
+## Decision
+
+`dsh-subagent` owns one canonical selection rule in `src/assistant-output.ts`: select the last non-empty assistant message; without one, select the accumulated `text-delta` stream; ignore empty-content messages. The incremental `AssistantOutputFold` implements the rule through `push(event)` for session-event transports, `pushText(text)` for chunk-only transports, and `collect()` for selection. `finalAssistantOutput(events)` applies it to a complete event suffix for the in-process `readResult` and Activation capture. The SDK backend folds notification events; the ACP backend exposes no complete assistant messages and folds raw chunk text. `SubagentResult.output` defines the result contract, and `subagent/end.lastAssistantMessage` uses the same rule. When a child produces neither form of output, the lifecycle field is absent rather than an empty array for both one-shot and continuable runs. A `max-tokens` or `aborted` result retains its actual stop reason.
+
+The foreground delegation tool uses the same selection. A non-`completed` result remains an `isError` tool result, but its message appends the child's partial text after the stop-reason headline so the parent model receives both the failure and available output.
+
+## Verification
+
+The keyless SDK backend test uses `FAKE_EMPTY_MESSAGE` to emit a usage-only terminal message. The `subagent-max-tokens-partial` ACP snapshot records a child that streams text and a tool call, ends at a tool-only max-tokens step with an empty usage message in its durable log, and returns the partial text through the parent's errored tool result. Unit coverage checks empty terminal messages, cancellation, message ordering, textless non-empty messages, and exclusion of tool-result content.
+
+## Alternatives considered
+
+**Fix each consumer in place without a shared helper.** Rejected: three independent selections had diverged, while observers of one run must agree on its output.
+
+**Stop the loop from appending the empty message.** Rejected: the message records usage and preserves the step in the durable log ("model-visible ⟺ logged"); changing session events to address output selection would affect every replay and projection consumer.
+
+**Treat empty-content messages as an error.** Rejected: the streamed text is the child's real partial answer, and the stop reason already tells the consumer the turn was cut short.
+
+## Consequences
+
+Multi-step children cut off at max-tokens report their earlier text; cancelled in-process children retain text streamed before the abort; one-shot and continuable `subagent/end` events agree with `SubagentResult.output`. A message whose content is non-empty but textless, such as reasoning-only content, is selected instead of streamed text because the rule tests content length rather than text presence. A non-empty message is also selected instead of text streamed after it: a child cancelled while streaming a later step reports its earlier complete message, while the stop reason records the truncation.
@@ -0,0 +1,31 @@
+# Agent Note: 用同一条选取规则在空终止消息后保留子代理输出
+
+Status: implemented
+
+[English](2026-08-10-subagent-empty-terminal-message-output.md) | 中文
+
+## 问题
+
+当 `max-tokens` 步骤只组装了工具调用块时，agent loop（智能体循环）会追加一条空内容的 `assistant/message`，因为 `BlockAssembler.blocks()` 会丢弃被截断的工具调用；这条消息仅记录 usage。三个消费方独立选取子 agent 的输出，并把这条 usage 记录当成输出。进程内驱动的 `readResult` 与 continuable Activation 的 `subagent/end` capture 不加过滤地选取最后一条 `assistant/message`，SDK 后端的观察器则让任何 `assistant/message` 优先于累积的文本。在被 max-tokens 截断的多步轮次中，最后那条空消息导致 `SubagentResult.output`、工具结果、遥测与 `subagent/end.lastAssistantMessage` 都漏掉真实的部分回答。进程内驱动也没有流式文本兜底，因此被取消的子 agent 若其唯一文本只存在于 `assistant/chunk` 事件中，也会报告 `[]`。
+
+## 决策
+
+`dsh-subagent` 在 `src/assistant-output.ts` 中拥有唯一的规范选取规则：选取最后一条非空 assistant 消息；没有时选取累积的 `text-delta` 流；忽略空内容消息。增量的 `AssistantOutputFold` 通过 `push(event)` 处理会话事件传输，通过 `pushText(text)` 处理仅分片传输，并通过 `collect()` 完成选取。`finalAssistantOutput(events)` 把规则应用于完整的事件后缀，供进程内 `readResult` 与 Activation capture 使用。SDK 后端折叠通知事件；ACP 后端不暴露完整的 assistant 消息，而是折叠原始分片文本。`SubagentResult.output` 定义结果约定，`subagent/end.lastAssistantMessage` 使用同一规则。子 agent 不产生这两种输出中的任何一种时，一次性与 continuable 运行的生命周期字段都会缺省，而不是空数组。`max-tokens` 或 `aborted` 结果保留实际的终止原因。
+
+前台委派工具使用同一选取规则。非 `completed` 的结果仍是 `isError` 工具结果，但其消息会在终止原因标题之后附上子 agent 的部分文本，让父模型同时接收失败信息与已有输出。
+
+## 验证
+
+无密钥 SDK 后端测试使用 `FAKE_EMPTY_MESSAGE` 发出一条仅记录 usage 的终止消息。`subagent-max-tokens-partial` ACP 快照记录一个子 agent：它流式输出文本与一次工具调用，结束于仅含工具调用的 max-tokens 步骤，持久化日志中含一条空的 usage 消息，并通过父侧的错误工具结果返回部分文本。单元覆盖检查空终止消息、取消、消息顺序、不含文本的非空消息，以及排除工具结果内容。
+
+## 考虑过的替代方案
+
+**各消费方就地修复、不抽共享辅助函数。** 之所以否决：三处独立选取已发生分歧，而同一次运行的观察方必须对其输出达成一致。
+
+**让 loop 不再追加空消息。** 之所以否决：这条消息记录 usage，并在持久化日志中保留该步骤（"model-visible ⟺ logged"）；为处理输出选取而改动会话事件，会影响所有 replay 与 projection 消费方。
+
+**把空内容消息视为错误。** 之所以否决：流式文本才是子代理真实的部分回答，且终止原因已经告诉消费方轮次被截断。
+
+## 后果
+
+被 max-tokens 截断的多步子 agent 会报告其更早的文本；被取消的进程内子 agent 保留中止前已流式的文本；一次性与 continuable 的 `subagent/end` 事件同 `SubagentResult.output` 一致。内容非空但不含文本的消息（例如仅含 reasoning 的内容）仍然优先于流式文本，因为规则检查内容长度，而不是文本是否存在。非空消息同样优先于其后才流式出的文本：子 agent 在流式输出后续步骤时被取消，报告的是更早那条完整消息，终止原因则记录该截断。
@@ -1,6 +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/2026-08-08-trusted-repository-package-code.md
-2026-08-08-trusted-repository-package-code.md: 387479b3b36a8bc5e145641ae40802b3090ced70
-2026-08-08-trusted-repository-package-code.zh.md: ecc325c3dd0a9e823f1411c3a809c30ada486289
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-08-11-preset-card-description-clamp.md
+2026-08-11-preset-card-description-clamp.md: 16ebf371d5af7c9e54fcc37819696b380856d5cb
+2026-08-11-preset-card-description-clamp.zh.md: 5b7a18f41e4b3e8acd681a001f6826b19ca7026d
@@ -0,0 +1,43 @@
+# Agent Note: Preset cards clamp their description instead of sizing the roster
+
+Status: implemented
+
+English | [中文](2026-08-11-preset-card-description-clamp.zh.md)
+
+## Problem
+
+A preset publishes its own `description`, of any length, and the settings section renders the roster as a card grid. The description had a `min-height` and no upper bound, while the grid sizes rows with `grid-auto-rows: 1fr` — which makes every implicit row the same height, not just the row holding the tall card. One long description therefore set the height of the whole roster: with a 250-character description in the custom group, all four cards measured 421px and the short-description cards filled with blank space.
+
+The description is also the field that tells presets apart, so hiding it is not an option; the card has to bound it and still make the whole text reachable.
+
+## Decision
+
+The description clamps to four lines and offers the rest through the shared `Tooltip`, attached only while the element actually overflows (`scrollHeight > clientHeight`, re-measured through a ResizeObserver because the settings pane width follows the window). This mirrors the chat stats line, which clamps to one line on the same measure-then-attach rule.
+
+Card height stays derived rather than fixed. With the description bounded, `grid-auto-rows: 1fr` already equalizes the grid, and a card carrying the broken-preset reason or a revealed path still sizes itself — a pixel height would clip both.
+
+Three smaller decisions ride along:
+
+- `.cardId` takes the card's free space with `margin-top: auto`, and the description no longer grows. A flex-stretched box leaves the clamp height and the box height disagreeing; sizing the clamped box by content alone keeps the behavior independent of that interaction.
+- The description carries `title=""`. An empty `title` means the element has no advisory information and the lookup stops there, so the card body's native tooltip does not climb to the description and a cut-off description answers with one bubble instead of two.
+- `Tooltip` gains an optional `maxWidth`. Its default half-viewport cap renders a description as a slab wider than the settings dialog it belongs to, spilling across the application behind it.
+- `Tooltip` also flips a `top` or `bottom` bubble to the other side when the viewport has no room for it, which its horizontal-only clamp previously left unhandled. Custom presets sit at the bottom of the roster and carry the longest descriptions, so the common case put a tall bubble under an anchor low on the page. The flip only moves into a side that genuinely fits, so an anchor with room on neither side keeps the requested placement rather than oscillating; sliding the bubble vertically instead would cover the text being read.
+
+A roster row that failed its shape check is badged `Failed to load` (`加载失败`) rather than `Broken` (`已损坏`). Discovery sets `broken` when the composition file is missing, unreadable, or malformed — most often a file the user just edited or deleted — so a damage claim overstates what was observed, and the verbatim reason under the badge already names the file and the fix.
+
+## Alternatives considered
+
+- **A fixed card height.** It states the intent directly but clips the two rows whose height legitimately varies: the broken-preset reason and the revealed preset directory.
+- **The native `title` attribute carrying the full description.** No measurement and no component, but a roughly one-second delay, operating-system styling, and it takes over the card's `set as default` hint across most of the card's area.
+- **Attaching the tooltip unconditionally.** It drops the ResizeObserver, at the cost of answering a hover over a short description with a bubble repeating what is already on the card.
+- **Expanding the clamp on hover.** It shows the text in place, and moves the grid under the pointer.
+
+## Consequences
+
+The section owns a small measured component and the shared primitive owns one more optional prop. In exchange, no card's height follows the longest description anywhere in the roster, and the whole description stays in the accessibility tree because the clamp is CSS rather than truncated text.
+
+The `title=""` suppression is pinned by a DOM assertion, not by observing the native tooltip: a browser tooltip is drawn outside the page and cannot be captured. If a browser ever resumes climbing past an empty `title`, the fallback is to drop the card body's `title` — its content is already in the body's `aria-label`.
+
+## Testing
+
+Package tests cover the three measurement outcomes (cut off, fitting, and a runtime without `ResizeObserver`) and the tooltip width cap. The web e2e goldens replay unchanged except `damaged.expected.md`, re-recorded for the badge copy.
@@ -0,0 +1,43 @@
+# Agent Note: 预设卡片截断自身描述，而不是由描述决定整份名单的高度
+
+Status: implemented
+
+[English](2026-08-11-preset-card-description-clamp.md) | 中文
+
+## 问题
+
+preset 自行发布 `description`，长度不限，而设置分区把名单渲染为卡片网格。描述只有 `min-height` 没有上限，网格则以 `grid-auto-rows: 1fr` 排布行——该取值让每一个隐式行等高，而不只是承载高卡片的那一行。因此一条长描述决定了整份名单的高度：自定义组里放入一条 250 字的描述后，四张卡片全部量得 421px，短描述卡片被大片空白填满。
+
+描述同时又是区分各个 preset 的字段，因此不能藏起来；卡片必须既给它设上限，又让全文仍然可达。
+
+## 决定
+
+描述截断为四行，其余内容通过共享的 `Tooltip` 呈现，且仅在元素确实溢出时才挂载（`scrollHeight > clientHeight`，并经 ResizeObserver 重新测量，因为设置面板宽度跟随窗口）。这与聊天统计行一致：它按同样的「先测量再挂载」规则截断为一行。
+
+卡片高度仍是推导得出而非固定。描述有了上限之后，`grid-auto-rows: 1fr` 本身就让网格等高，而承载损坏原因或已展示目录的卡片仍能按自身内容定高——写死像素高度会把两者一并裁掉。
+
+随之而来三个更小的决定：
+
+- `.cardId` 以 `margin-top: auto` 吃掉卡片的空余空间，描述不再拉伸。被 flex 拉伸的盒子会让截断高度与盒子高度不一致；让截断盒子只按内容定高，行为便不依赖这层交互。
+- 描述带有 `title=""`。空 `title` 表示该元素没有提示信息，查找就此停止，因此卡片主体的原生 tooltip 不会向上找到描述，被裁切的描述只回应一个气泡而不是两个。
+- `Tooltip` 新增可选的 `maxWidth`。它默认的半视口上限会把描述渲染成比所属设置弹窗还宽的一整块，溢出到背后的应用界面上。
+- `Tooltip` 同时在视口放不下时把 `top` 或 `bottom` 气泡翻到另一侧，此前它只做水平收敛。自定义 preset 位于名单末尾、又恰恰承载最长的描述，因此常见情形正是让一个高气泡挂在页面靠下的锚点之下。翻转只会移向确实放得下的一侧，两侧都放不下时保持请求的位置而不来回摆动；改为垂直滑动则会盖住正在阅读的文本。
+
+形状检查未通过的名单行，徽记从 `Broken`（`已损坏`）改为 `Failed to load`（`加载失败`）。discovery 在组装文件缺失、读不出或格式错误时置位 `broken`——最常见的是用户刚编辑或删除的文件——因此断言损坏超出了观察到的事实，而徽记下方原样展示的原因本就点名了文件与修法。
+
+## 备选方案
+
+- **写死卡片高度。** 它直接表达了意图，却会裁掉两处高度本就可变的行：损坏预设的原因行和已展示的预设目录。
+- **用原生 `title` 属性承载完整描述。** 无需测量也无需组件，代价是约一秒的延迟、操作系统的样式，以及在卡片大部分区域内顶替掉「设为默认」的提示。
+- **无条件挂载 tooltip。** 省掉 ResizeObserver，代价是把鼠标停在短描述上时，弹出一个重复卡片已有内容的气泡。
+- **hover 时展开截断。** 它就地展示文本，同时让网格在指针下方发生位移。
+
+## 后果
+
+分区多了一个带测量的小组件，共享基元多了一个可选 prop。换来的是：任何卡片的高度都不再跟随名单中最长的那条描述；而且截断由 CSS 完成而非截短文本，完整描述始终留在无障碍树中。
+
+`title=""` 的抑制作用由一条 DOM 断言钉住，而非通过观察原生 tooltip：浏览器 tooltip 画在页面之外，无法被捕获。若某个浏览器日后重新越过空 `title` 继续向上查找，退路是去掉卡片主体的 `title`——它的内容已经在主体的 `aria-label` 里。
+
+## 测试
+
+包内测试覆盖三种测量结果（被裁切、放得下、运行时没有 `ResizeObserver`）以及 tooltip 的宽度上限。web e2e golden 除 `damaged.expected.md` 按徽记文案重录外，其余原样回放通过。
@@ -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/2026-07-06-approval-seam.md
-2026-07-06-approval-seam.md: 7c830d93f19a40ab193cfebabca854882ab68d62
-2026-07-06-approval-seam.zh.md: 9dedfddadc23b0da44b28e8750508653ee20bb83
+2026-07-06-approval-seam.md: 8aa9986139dae77e08c166b72545bfa688a389e0
+2026-07-06-approval-seam.zh.md: ef4ccf5fd2b54888a648737866ff6f5fe1678882
@@ -123,7 +123,7 @@ Costs and accepted limits:
 - **Who decides whether a call asks in the first place?** Policy producers: a hook returning `permissionDecision: ask`, any `tools/pre-execute` listener, or the sandbox escalation gate. The seam and the bridge only route and answer; neither injects its own judgment about what deserves a prompt.
 - **What happens when the user dismisses the prompt, or the turn aborts mid-ask?** Dismissal maps to `cancelled` with its own deny text. An already-aborted signal settles `cancelled` without dispatching; an abort during the ask discards the late answer. When both audit appends commit, either path records one pair, never two.
 - **What if the client answers with an option the harness never offered?** Any selection other than the offered `allow_once` maps to `rejected` — an unknown optionId from a non-conforming client can never grant.
- **How do subagents' approvals route?** An agent no answerer owns delegates through the whole waterfall and fails closed — in-process subagents are deliberately unanswerable. A `'never'` parent seeds that override into each in-process child's log ([decision](2026-07-25-subagent-policy-inheritance.md)), so the child is told up front instead of asking into the empty waterfall. `subagent-acp`'s child-side auto-answer is separate; routing a child's asks to the parent controller is deferred (§ Deferred).
+- **How do subagents' approvals route?** They do not: delegation pins every in-process child to `'never'` ([approvals-pinned decision](2026-08-10-subagent-approval-pinned-never.md)), so each child ask resolves `rejected` before any answerer and the child is told up front through its runtime context. `subagent-acp`'s child-side auto-answer is separate; routing a child's asks to the parent controller is deferred (§ Deferred).
 - **What does `policy: 'never'` actually change at runtime?** The service resolves every ask for that session to `rejected` before dispatching any answerer (in-service, so no registration order can bypass it); the next atomic runtime-context snapshot states the policy; each successful auto-rejection records the audit pair.
 - **What happens across a hot reload, or when an answerer unloads mid-session?** Answerers dispose with their owning fiber, so the next ask degrades to `unavailable` instead of hanging on a dead channel; remounting re-registers the answerer with no catch-up state.
 - **Where does a client get approval context?** The request carries the exact `callId` and the asker's human-readable `reason`; channel adapters may correlate richer tool-call state without duplicating arguments in the approval seam.
@@ -123,7 +123,7 @@ ACP 桥只应答其会话映射所拥有的精确 agent 对象。它携带既有
 - **谁决定一次调用是否需要 ask？** 策略生产者：返回 `permissionDecision: ask` 的钩子、任何 `tools/pre-execute` 监听器、或沙箱升级门禁。seam 和桥只负责路由和应答；二者都不注入自己对「什么值得弹出提示」的判断。
 - **用户关闭提示或轮次在 ask 进行中中止时会发生什么？** 关闭映射为 `cancelled` 并携带自己的拒绝文本。已中止的 signal 直接结算为 `cancelled` 而不派发；ask 进行中的中止丢弃迟到的应答。当两个审计追加都提交时，任一路径都记录恰好一对事件，绝不会两对。
 - **如果客户端以 harness 从未提供的选项应答呢？** 除已提供的 `allow_once` 之外的任何选项都映射为 `rejected`——来自不合规客户端的未知 optionId 永远不能授权。
- **subagent 的审批如何路由？** 没有应答者拥有的 agent 穿过整个 waterfall 委派并失败关闭——进程内 subagent 被刻意设计为不可应答。`'never'` 父级会把该覆盖项预置到每个进程内子 agent 的日志中（[决策](2026-07-25-subagent-policy-inheritance.md)），因此子 agent 一开始就会得知，而不是向空的 waterfall 发出 ask。`subagent-acp` 的子侧自动应答是独立的；将子 agent 的 ask 路由到父控制器已延后（§ 延后）。
+- **subagent 的审批如何路由？** 不路由：委派会把每个进程内子 agent 钉定为 `'never'`（[审批钉定决策](2026-08-10-subagent-approval-pinned-never.md)），因此子 agent 的每次 ask 都在任何应答者之前解析为 `rejected`，子 agent 则通过其运行时上下文一开始就会得知。`subagent-acp` 的子侧自动应答是独立的；将子 agent 的 ask 路由到父控制器已延后（§ 延后）。
 - **`policy: 'never'` 在运行时实际改变了什么？** 服务在派发任何应答者之前，将该会话的每次 ask 解析为 `rejected`（在服务内部，因此没有注册顺序能绕过它）；下一份原子化的运行时上下文快照会声明该策略；每次成功的自动拒绝都会记录审计对。
 - **热重载或应答者在会话中途卸载时会发生什么？** 应答者随其拥有的 fiber 一起 dispose，因此下一次 ask 降级为 `unavailable` 而非挂在死通道上；重新挂载会重新注册应答者，无需追赶状态。
 - **客户端从哪里获得审批上下文？** 请求携带精确的 `callId` 和发起方的人类可读 `reason`；通道适配器可自行关联更丰富的工具调用状态，而无需在审批 seam 中重复携带参数。
@@ -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/2026-07-06-sandbox.md
-2026-07-06-sandbox.md: 0214a202a41983c76fa25e3a82e1cfaec70a0d55
-2026-07-06-sandbox.zh.md: bd8dcdb74f723a955ea2a1cc5b224ef2ded4a8d5
+2026-07-06-sandbox.md: e1ec35cdcf3be2af03232f823d9a9cee57b4e8e8
+2026-07-06-sandbox.zh.md: d3b74d441f72fc4c218979c327546031fd1b24f1
@@ -171,7 +171,7 @@ Costs and accepted limits:

 - **The one-wrapper illusion is given up knowingly.** A `tools/pre-execute` wrapper plus prompt conventions does not solve sandbox approval — the correct design costs structured denials, native runner probes, per-call policy carriage, and consistent cross-family enforcement, and this design pays it.
 - **`read-only` became a cross-family boundary through a follow-up.** This RFC shipped bash-only enforcement; the [cross-family fs sandbox RFC](2026-07-14-cross-family-fs-sandbox.md) extends the same mode vocabulary to the filesystem tools through a sandboxed `ctx.fs` provider and relocates the mode/root config and the `sandbox/mode` override to `ctx.sandboxPolicy` (§ In-process tools).
- **Windows has no backend.** Its chain slot is reserved empty — fail-closed, never a fallthrough; filling it is a deferred phase.
+- **Windows is a partial backend.** This RFC originally reserved an empty, fail-closed win32 chain; the later [Windows ACL sandbox decision](2026-08-08-windows-acl-restricted-token-sandbox.md) filled it with the restricted-token runner. Its Everyone and hard-link gaps are reported as `enforcement: 'partial'`, never promoted to the full promise.
 - **The Seatbelt rung leans on Apple's deprecated-but-shipped `sandbox-exec` CLI.** As darwin's sole candidate it is selected without probing, so a future removal under a usable workdir surfaces as a runner-attributable spawn failure and an executable refusal through its fatal signature — both become `SANDBOX_UNAVAILABLE`, and the command never runs; fail closed, never open.
 - **Landlock confinement is only as complete as the running kernel's ABI.** Reported as `enforcement: 'partial'` rather than refused — the deliberate trade that keeps the fallback available on older-kernel hosts.
 - **Runner attribution uses an in-band protocol.** Exit status plus stderr cannot cryptographically identify the writer, so a confined child can mimic a fatal runner line and status to cause an availability/diagnostic false attribution. The conjunction and exact notice exclusion reduce accidental matches; this is not a sandbox bypass because the child is already confined.
@@ -171,7 +171,7 @@ fs/web/todo 在进程内执行，因此它们的沙箱语义是各自能力边

 - **单一包装的幻觉被有意放弃。**`tools/pre-execute` 包装加提示词约定无法解决沙箱批准——正确的设计需要结构化拒绝、原生 runner 探测、按调用策略承载和一致的跨工具族强制，本设计为此付出了代价。
 - **`read-only` 通过后续设计成为跨工具族边界。** 本 Agent Note 最初只交付 bash 强制；[跨工具族 fs 沙箱 Agent Note](2026-07-14-cross-family-fs-sandbox.md) 通过沙箱化的 `ctx.fs` 提供方把同一模式词汇扩展到文件系统工具，并将 mode/root 配置和 `sandbox/mode` 覆盖迁移到 `ctx.sandboxPolicy`（§ 进程内工具）。
- **Windows 没有后端。** 其链槽保留为空——失败关闭，绝不穿透；填充它是延迟阶段。
+- **Windows 后端只提供部分强制执行。** 本 RFC 最初预留了一条空的、失败关闭的 win32 链；后续的 [Windows ACL 沙箱决策](2026-08-08-windows-acl-restricted-token-sandbox.md)以受限令牌 runner 填充了它。其 Everyone 与硬链接缺口报告为 `enforcement: 'partial'`，绝不提升为完整承诺。
 - **Seatbelt 层级依赖 Apple 已弃用但仍交付的 `sandbox-exec` CLI。** 作为 darwin 的唯一候选，它无需探测即被选中，因此在 workdir 可用时，未来移除会表现为可归因于 runner 的 spawn 失败，可执行文件拒绝则通过其致命签名体现——两者都会变为 `SANDBOX_UNAVAILABLE`，且命令绝不会运行；失败关闭，绝不开放。
 - **Landlock 约束的完整度取决于运行内核的 ABI。** 报告为 `enforcement: 'partial'` 而非拒绝——这是有意的权衡，使备选在旧内核主机上仍可用。
 - **Runner 归因使用带内协议。** 退出状态与 stderr 无法以密码学方式识别写入者，因此受限子进程可以模仿 runner 的致命诊断行和状态，造成可用性或诊断误归因。多项证据的合取与精确通知排除减少了意外匹配；这不是沙箱绕过，因为子进程已经受到限制。
@@ -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/2026-07-07-mcp-client-plugin.md
-2026-07-07-mcp-client-plugin.md: 756a5c4dc9f1152ecb1955d93b8dc47fcf07c661
-2026-07-07-mcp-client-plugin.zh.md: 9f45203c479379087c5a19eccce3ab7a339d4ed0
+2026-07-07-mcp-client-plugin.md: 077d978d8815d759574f89ed524ab3bc8c24a267
+2026-07-07-mcp-client-plugin.zh.md: 8f58c9359ca8447717cc353f97f1333370fa6fda
--- a/Show More
+++ b/Show More