[objectstack-fleet]

Requested by 田騏 · project thread

Part of #30512
Fixes #28107
Clause-②: no
Metadata-protocol: none

维护者速读

• 改了什么：给 @objectstack/platform-objects 加上法语（fr）：平台对象、Studio 的元数据表单、Setup/Studio/Account 导航、系统概览仪表板和用户/组织记录页的内嵌文字，做法与刚合并的 ar（i18n(platform-objects): ship the platform objects, Setup, Studio and Account in Arabic #31120）逐文件一致。
• 为什么改：语言设为 fr 的用户在主页动态里读到「A modifié User « … » : Banned et Ban Reason」，法语句子里夹着英文字段名（i18n: French is not a platform locale, so a user whose locale is fr reads English sys_user labels (Home feed field labels) #28107）；fr 也是 i18n phase 3 (platform metadata and built-in emails): extraction and bundle checks read PLATFORM_LOCALES; backfill the seven missing locales #30512 要补的第三个语言。
• 风险与回滚：只新增 fr 文件、注册行和测试里的语言列表，不改任何已有语言的文字；新文字同样走 source-hash 回退。回滚即 revert 本 PR。
• 先看：packages/platform-objects/src/apps/translations/fr.ts（导航钉住项），packages/plugins/plugin-audit/src/activity-summary-fr-sys-user-labels.test.ts（i18n: French is not a platform locale, so a user whose locale is fr reads English sys_user labels (Home feed field labels) #28107 的守护测试），fr.metadata-forms.generated.ts 里 repeater 行属性那几处带上父名的译法。
• 两分钟验证：cd packages/plugins/plugin-audit && npx vitest run src/activity-summary-fr-sys-user-labels.test.ts，预期 5 个测试通过，句子为「A modifié Utilisateur « Carl Edited » : Banni et Motif du bannissement」。
• 需要你做的：生成文件让 PR 超过 3000 行，座位契约复审 ACCEPT 后请人工合并。

What changes

• New fr bundles in packages/platform-objects/src/apps/translations: fr.objects.generated.ts (1697 leaves), fr.metadata-forms.generated.ts (1263 leaves), their provenance table fr.source-hashes.generated.ts (183 deliberate byte copies of en, listed under Acceptance notes), the hand-authored fr.ts (140 leaves) and fr.source-hashes.ts recording the en digest of every hand leaf.
• Registration: the extract config's --locales docstring and translations: baseline, setup.translation.ts, the apps barrel, and metadata-translations/index.ts.
• Inline maps: 36 in sys-user.page.ts, 6 in sys-organization.page.ts, and the System Overview null label in system.datasets.ts.
• Wording pins: nav_account_approvals reads as plugin-approvals' my_pending label (Mes approbations en attente), nav_assignment_rules as sys_assignment_rule's plural label (Règles d’attribution), each nav_settings_* (including nav_settings_access_policy) as its service-settings hub title, nav_settings_hub as the console's backToHub (Tous les paramètres), and the four feat(console): put Connect an AI Assistant where each person looks, and call it that #30935 keys: grp_account_ai_assistant = Assistant IA; both nav_connect_agent entries and pages.connect_agent.label = Connecter un assistant IA.
• fr.source-hashes.ts also records nav_settings_access_policy, which the ar and zh-Hant tables leave legacy-trusted.
• Changeset .changeset/30512-platform-objects-fr.md (patch).

Tests

• fr added to the locale lists of every test i18n(platform-objects): ship the platform objects, Setup, Studio and Account in Arabic #31120 widened for ar (vocabulary, nav parity, dead-key tombstone, source-hash, copy tests, record-page tabs, plugin translation loads), plus plugin-approvals approval-status-vocabulary, plugin-security people-org-assignment-nav-contribution and service-settings setup-nav-hub-parity. The cli platform-page-i18n-parity LOCALE_KEY shape already accepts fr.
• New packages/plugins/plugin-audit/src/activity-summary-fr-sys-user-labels.test.ts (i18n: French is not a platform locale, so a user whose locale is fr reads English sys_user labels (Home feed field labels) #28107): an i18n service holding what a booted kernel holds (this plugin's bundle and the platform Setup bundle, every locale) composes a sys_user update summary for a reader whose locale is fr or fr-FR. The reader gets A modifié Utilisateur « Carl Edited » : Banni et Motif du bannissement, and a tracked change names Motif du bannissement; an en control keeps Updated User "Carl Edited": Banned and Ban Reason. Ablation: with the platform fr bundle left out of the service, three cases fail with A modifié User « Carl Edited » : Banned et Ban Reason, the sentence i18n: French is not a platform locale, so a user whose locale is fr reads English sys_user labels (Home feed field labels) #28107 reported.

Acceptance notes

• Every leaf is translated. The 183 byte copies of en the extract records are French words spelled like the English (Description, Type, Source, Version, Navigation, Configuration, Kanban, …), identifiers and header names (ID, API, JWKS, JWKS URI, Message-ID, Reply-To, Cc, cron, snake_case), lowercase option values that read the same in French (marketing, notification, workflow, package), and the five JSON or scope examples kept verbatim as every locale keeps them.
• Where a catalog test refuses an en echo (repeater row properties, dashboard header children, System Overview captions), the leaf names its parent instead (ID du nœud, Description de l’option, Type de variable, Boutons d’action, Nombre de sessions), as es-ES does.
• Typography follows ui/packages/i18n fr: a plain space before : ; ? !, plain spaces inside « », the typographic apostrophe, sentence case.
• Position on the add-record button is Position (the place); the job position is Poste. A hook's Body is Corps. Adjectives agree with the object they qualify (Toutes, Révoquées, Désactivée, …).
• Business units are Unités commerciales; no sys_business_unit* label says Département (bu-rename-consistency).

Verification

• os i18n extract packages/platform-objects/scripts/i18n-extract.config.ts --locales=zh-CN,ja-JP,es-ES,zh-Hant,ar,fr --fill=default --source-hashes: regenerates the two fr bundles byte-identical and leaves every other locale unchanged, before and after the merge of origin/main.
• vitest run in platform-objects: 119 files, 1746 tests passed. plugin-approvals approval-status-vocabulary (6), plugin-security people-org-assignment-nav-contribution (5), service-settings setup-nav-hub-parity (15), cli platform-page-i18n-parity (27) and the new plugin-audit test (5) passed.
• platform-objects typecheck: exit 0.
• dispatch-gates --ran at dd9303a (after the merge of origin/main): 64 derived families, 64 run, 0 NOT-MEASURED, all exit 0 (including check:i18n, check:dts-closure, check:dual-build-cjs-loads).
• pnpm lint: exit 0. "Lint · docs & code guards" lane: 52 of its 53 run steps exit 0 locally (including check:i18n-stale-fill and check:cli-command-ids); the 53rd is lane-verdict.mjs, which reads GITHUB_JOB and only measures in CI.

🤖 Generated with Claude Code

[objectstack-fleet]

domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx

Contract review — ACCEPT

Head-sha: dd9303ac18aca10d6db917441390352eabfdab38

ACCEPT · default tier, in-seat · 2026-11-07T19:06Z

This PR touches none of the four contract faces (Clause-②: no, Metadata-protocol: none, no packages/spec/src, no governed text), so it gets a default-tier review in-seat with no second agent. It is the model #31120 (ar) repeated for fr, and it carries Fixes #28107.

• Coverage: 3100 en leaves and 3100 fr leaves, 0 untranslated (objects 1697, metadata forms 1263, hand-authored 140). All 43 inline maps are filled. Re-extracting with …,ar,fr reproduces both generated bundles byte for byte. The file list matches i18n(platform-objects): ship the platform objects, Setup, Studio and Account in Arabic #31120's (eb68b98149).
• Spot-read of fr.ts: the Setup nav reads idiomatically and consistently (Règles d’attribution, Centre d’approbations, Politique d’accès, Journal d’audit). It uses typographic apostrophes, and each nav_settings_* entry equals its service-settings hub title, which the hub-parity test pins.
• Your questions:
  1. The 183 recorded echoes are accepted: French cognates (Description, Type, Version), identifiers, lowercase option values and JSON examples. Where a catalog test refuses an echo, the parent-qualified label follows the es-ES precedent.
  2. The digest for nav_settings_access_policy in fr.source-hashes.ts is accepted. The header says a recorded digest is strictly better than legacy trust.
  3. The i18n: French is not a platform locale, so a user whose locale is fr reads English sys_user labels (Home feed field labels) #28107 test in plugin-audit is accepted. That package composes the Home feed sentence, the claim names that test, and its ablation (no platform fr bundle) reproduces the card's English labels.
• Gates: dispatch-gates --ran 64/64 exit 0, 0 NOT MEASURED. pnpm lint 0. Docs & code guards 52/53, the remaining one being lane-verdict, which only CI can run. CI on this head is still running.
• The changeset is @objectstack/platform-objects patch with Clause-②: no / Metadata-protocol: none. plugin-audit changes only in a test, so it needs no changeset.

Landing: at 7535 changed lines the size gate applies, so os-zhuang or hotlong must approve before the queue. The PR is labelled pm:awaiting-merge and assigned. Before it is enqueued, the card thread merges main again and reruns check:app-nav-i18n, hub parity and platform-page-i18n-parity on the merged tree, because ar was dequeued once for strings main added after its branch point.

[github-actions]

📓 Docs Drift Check

This PR changes 1 package(s): @objectstack/platform-objects, touching 291 documentable anchor(s). ⚠️ 1 changed file(s) yielded no anchor (packages/platform-objects/src/apps/translations/index.ts), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

66 hand-written doc(s) name something this change touched — list omitted above 15 rows. Re-derive on the tree named below: node scripts/docs-audit/affected-docs.mjs --json 57675a21470626b71f086d3fba0147a9440a3e03.

⛔ 19 release-owned page(s) also affected — read-only, see AGENTS.md Documentation Guardrails.

What this run could not see

• 1 changed file(s) yielded no anchor (packages/platform-objects/src/apps/translations/index.ts) — pages documenting those are invisible to this run
• 2 anchor(s) matched too much of the corpus to be a work list: sys_user (symbol, 49 pages), snake_case (literal, 52 pages)
• 33 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 63 of 226 client-bound route-ledger rows — the other 163 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 163: 0 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 59 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 104 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.
• a key NAME is not a key, so the hand re-read the line above prescribes can land on the wrong schema. The same spelling is authorable on one governed type and a [REMOVED] tombstone on another for each of active, aria, joins, objects, template, tools and version (censused on [finding] tools is a key on BOTH AgentSchema (tombstoned, dead) and SkillSchema (live, cloud-attested), so a name-based search attributes skill examples to the agent key — it produced a false stop-the-line alarm on PR #19059 #19093 over the liveness ledger's governed types, top-level keys); nothing in a search result distinguishes the two, so a grep hit on a LIVE example reads as evidence about the DEAD key. Measured on fix(spec): the agent.tools liveness row says dead — it claimed live on a key the schema tombstoned #19059: content/docs/ai/agents.mdx was reported as contradicting the agent.tools tombstone over its tools: example at :161, which is inside the defineSkill({ block opened at :155 — the page was already correct. Settle ownership by PARSING the value against both schemas, never by the name: that literal PASSES SkillSchema, and as an AgentSchema it FAILS at tools with the tombstone prescription. ⛔ These names are not the whole class — a key retired through a .strict() guidance map leaves no tombstone in the walked shape and none of them here (tool.category, live as AIToolDefinition.category).

Coarse fallback — 3 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 57675a21470626b71f086d3fba0147a9440a3e03 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from c97f9713256fdabed47395f19aee4d955fa54061 — the merge of head e707147883c0a4f5b186eae38b7bba0e21a1a89c into base 57675a21470626b71f086d3fba0147a9440a3e03, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin c97f9713256fdabed47395f19aee4d955fa54061 && git checkout c97f9713256fdabed47395f19aee4d955fa54061
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 57675a21470626b71f086d3fba0147a9440a3e03 e707147883c0a4f5b186eae38b7bba0e21a1a89c && git checkout -B drift-repro 57675a21470626b71f086d3fba0147a9440a3e03 && git merge --no-ff e707147883c0a4f5b186eae38b7bba0e21a1a89c

node scripts/docs-audit/affected-docs.mjs --json 57675a21470626b71f086d3fba0147a9440a3e03

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

Advisory only, and a precision-first one (#9192): a page is listed because it names a
symbol, wire route or SDK method this diff touched — not because it mentions a changed
package. Each row says which anchor put it there, so a wrong row is reportable rather than
merely annoying. To re-verify, run the docs-accuracy-audit workflow scoped to these files:
node scripts/docs-audit/affected-docs.mjs 57675a21470626b71f086d3fba0147a9440a3e03 → pass the list as
args.docs, on the commit named under Which tree this was computed on.

[objectstack-fleet]

domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx

席位判读 — Test Core (8/8) 判为 flake,重跑一次 · 2026-11-07T19:18Z

• 红的是 run 42404452537 的 Test Core (8/8)(job 128499475343),只有 packages/plugins/plugin-auth/src/organization-access-policy.test.ts 一个用例(the absolute limit is judged too, and a target with no limits admits the switch,第 422 行):Expected admitted,Received ended。
• 本分支上的夹具把会话做成恰好 60 分钟,对一小时上限;判定在夹具之后读自己的时钟,只要过去一毫秒就算超限。这是时钟边界抖动,main 已由 test(plugin-auth): keep the absolute-limit switch fixture off the one-hour boundary #31213(df22adc370)把夹具改成 59 分钟修掉,本分支早于它。
• 本 PR 只改 fr i18n bundle 与注册,不碰 packages/plugins/plugin-auth/**。
• 判为 flake。席位按 platform-readings 规则用 rerun-failed.mjs 重跑一次失败作业;第二次红即按真红处理。获批后线程合并 main 时会一并带上 test(plugin-auth): keep the absolute-limit switch fixture off the one-hour boundary #31213。

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 42426481440 红了。队列跑的是全量套件（PR 侧 CI 只跑 affected 子集），
所以失败的测试可能在本 PR 没碰过的包里 —— 那不是重排能修的。每次盲目重排都会让排在后面的所有 PR 重建一轮。

分类：failure —— 按下面的日志分诊。

失败的 job（日志抽取，best effort）：

• Test Core (3/8) — 失败步骤: Run this shard's tests

  @objectstack/platform-objects:test:  FAIL  src/apps/translations/setup-nav-dead-key-tombstone.test.ts > conditionally-contributed Setup nav ids stay labelled (#6659) > fr carries a label for every con
    ↳ 失败原因: @objectstack/platform-objects:test: AssertionError: Setup nav ids with no label — no boot-time gate can see these; see this file header: expected [ 'nav_scim_credentials' ] to deeply equal []
  

• Test Core (4/8) — 失败步骤: Run this shard's tests

  @objectstack/cli:test:  FAIL   integration  src/commands/migrate/plan.telemetry-sibling.integration.test.ts > os migrate plan / apply plan each object against the database it lives in (#22579) > the c
    ↳ 失败原因: @objectstack/cli:test: AssertionError: a table the serving boot did not provision
  

↳ 失败原因 是判读的关键：超时（Test timed out in … / Hook timed out in …）多半是负载/时序，不是本 PR 的回归；
断言（AssertionError: …）才指向真实的行为改变。两者的 FAIL 行长得一模一样，只有这一行能区分。

⚠️ 断言这一侧有一类例外，判据是断言在测什么，不是它是不是 AssertionError。 断言的对象是产品行为（一个值、一个形状、一次拒收）⇒ 照上面读：真实的行为改变，去查，⛔ 不要重排掉；
断言的对象是这次实验自身的有效性前提（跑完的耗时、负载下的先后、任何只在时间预算内才成立的条件）⇒ 它跟超时是同一类，同样对负载敏感，重排一次是合法的判别手段。
识别是机械的：断言的消息或它比较的值本身点名了一段时长、一个时间戳、一个耗时计数。实测过的一对 —— AssertionError: SecurityPlugin.init() ran: expected false to be true 测的是产品行为（真回归）；
AssertionError: this run took over a second, so second-precision stamps could have differed too: expected 1006 to be less than 1000 测的是实验前提：它守护的那条不变式当时是绿的，同一个 head 原样重排一次即成功。
穿着 AssertionError 外衣的时间测量，仍然是时间测量。（⛔ 这只改「怎么读一次红」，不改「哪些测试可以重排」——后者由别处管。）

跨 PR 相同签名（24h，按失败测试文件聚合）：

• src/apps/translations/setup-nav-dead-key-tombstone.test.ts — 24h 窗口内只有本 PR 撞到过，暂不汇总（再有一个不同 PR 撞到就会自动开汇总 issue）。
• src/commands/migrate/plan.telemetry-sibling.integration.test.ts — 24h 窗口内只有本 PR 撞到过，暂不汇总（再有一个不同 PR 撞到就会自动开汇总 issue）。
• ⚠️ 24h 评论账本没读完（超过 5 页仍未读到窗口尽头），所以上面的「不同 PR 数」是下界，不是全量。

历史信号：

• 本 PR 过去 24h 无队列失败记录（首次）。
• 过去 24h 队列共有 21 个失败构建（不含本次）。

分诊清单：

1. 失败测试在本 PR 改动的包里 → 真回归，修 PR。
2. 失败测试与本 PR 无关 → 看上面的「跨 PR 相同签名」；已有汇总 issue ⇒ flaky/环境问题实锤，去那张 issue 上谈，修好前重排只会再烧一轮全队列。
3. 两者都不是 → 可能与同组 PR 语义冲突；等前面的 PR 落地或失败出队后再重排一次即可，不要连续重排。

---

Generated by Claude Code · merge-queue-triage workflow (#4859)

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 42427285234 红了。队列跑的是全量套件（PR 侧 CI 只跑 affected 子集），
所以失败的测试可能在本 PR 没碰过的包里 —— 那不是重排能修的。每次盲目重排都会让排在后面的所有 PR 重建一轮。

分类：failure —— 按下面的日志分诊。

失败的 job（日志抽取，best effort）：

• Test Core (3/8) — 失败步骤: Run this shard's tests

  @objectstack/platform-objects:test:  FAIL  src/apps/translations/setup-nav-dead-key-tombstone.test.ts > conditionally-contributed Setup nav ids stay labelled (#6659) > fr carries a label for every con
    ↳ 失败原因: @objectstack/platform-objects:test: AssertionError: Setup nav ids with no label — no boot-time gate can see these; see this file header: expected [ 'nav_scim_credentials' ] to deeply equal []
  

• Test Core (8/8) — 失败步骤: Run this shard's tests

  @objectstack/example-todo:test:  FAIL  test/loop-body-iteration.test.ts > #19206 — the loop body runs and binds `currentTask` > ESCALATION: the run reaches `notify_owner` and the notice renders the re
    ↳ 失败原因: @objectstack/example-todo:test: Error: Test timed out in 5000ms.
  @objectstack/example-todo:test:  FAIL  test/loop-body-iteration.test.ts > #19206 — the loop body runs and binds `currentTask` > EVERY item is processed, and each one carries its own values
    ↳ 失败原因: @objectstack/example-todo:test: Error: Test timed out in 5000ms.
  

↳ 失败原因 是判读的关键：超时（Test timed out in … / Hook timed out in …）多半是负载/时序，不是本 PR 的回归；
断言（AssertionError: …）才指向真实的行为改变。两者的 FAIL 行长得一模一样，只有这一行能区分。

⚠️ 断言这一侧有一类例外，判据是断言在测什么，不是它是不是 AssertionError。 断言的对象是产品行为（一个值、一个形状、一次拒收）⇒ 照上面读：真实的行为改变，去查，⛔ 不要重排掉；
断言的对象是这次实验自身的有效性前提（跑完的耗时、负载下的先后、任何只在时间预算内才成立的条件）⇒ 它跟超时是同一类，同样对负载敏感，重排一次是合法的判别手段。
识别是机械的：断言的消息或它比较的值本身点名了一段时长、一个时间戳、一个耗时计数。实测过的一对 —— AssertionError: SecurityPlugin.init() ran: expected false to be true 测的是产品行为（真回归）；
AssertionError: this run took over a second, so second-precision stamps could have differed too: expected 1006 to be less than 1000 测的是实验前提：它守护的那条不变式当时是绿的，同一个 head 原样重排一次即成功。
穿着 AssertionError 外衣的时间测量，仍然是时间测量。（⛔ 这只改「怎么读一次红」，不改「哪些测试可以重排」——后者由别处管。）

跨 PR 相同签名（24h，按失败测试文件聚合）：

• test/loop-body-iteration.test.ts — 24h 内已弹出 4 个 PR（i18n(platform-objects): ship the platform objects, Setup, Studio and Account in French #31233、fix(auth): record an MCP access token as used only after the MCP door admits the call #31247、feat(plugin-auth, plugin-audit): a sign-in the IP allow-list refuses writes one login_failed row, bounded per address #31249、feat(plugin-auth)!: an organization administrator reads their judged client address; the console's retired MFA notices are gone #31258），分属 3 条投机栈 ⇒ 3 次独立命中。汇总 issue：Queue-flake anchor: test/loop-body-iteration.test.ts #31294（已刷新）
• src/apps/translations/setup-nav-dead-key-tombstone.test.ts — 24h 窗口内只有本 PR 撞到过，暂不汇总（再有一个不同 PR 撞到就会自动开汇总 issue）。
• ⚠️ 24h 评论账本没读完（超过 5 页仍未读到窗口尽头），所以上面的「不同 PR 数」是下界，不是全量。

历史信号：

• ⚠️ 本 PR 过去 24h 已在队列失败 1 次（不含本次）。 内容未变而反复失败 ⇒ 高度怀疑 flaky 测试或与同组 PR 的语义冲突，重排不解决。
• 过去 24h 队列共有 22 个失败构建（不含本次）。

分诊清单：

1. 失败测试在本 PR 改动的包里 → 真回归，修 PR。
2. 失败测试与本 PR 无关 → 看上面的「跨 PR 相同签名」；已有汇总 issue ⇒ flaky/环境问题实锤，去那张 issue 上谈，修好前重排只会再烧一轮全队列。
3. 两者都不是 → 可能与同组 PR 语义冲突；等前面的 PR 落地或失败出队后再重排一次即可，不要连续重排。

---

Generated by Claude Code · merge-queue-triage workflow (#4859)

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 42429366760 红了。队列跑的是全量套件（PR 侧 CI 只跑 affected 子集），
所以失败的测试可能在本 PR 没碰过的包里 —— 那不是重排能修的。每次盲目重排都会让排在后面的所有 PR 重建一轮。

分类：failure —— 按下面的日志分诊。

失败的 job（日志抽取，best effort）：

• Test Core (3/8) — 失败步骤: Run this shard's tests

  @objectstack/platform-objects:test:  FAIL  src/apps/translations/setup-nav-dead-key-tombstone.test.ts > conditionally-contributed Setup nav ids stay labelled (#6659) > fr carries a label for every con
    ↳ 失败原因: @objectstack/platform-objects:test: AssertionError: Setup nav ids with no label — no boot-time gate can see these; see this file header: expected [ 'nav_scim_credentials' ] to deeply equal []
  

↳ 失败原因 是判读的关键：超时（Test timed out in … / Hook timed out in …）多半是负载/时序，不是本 PR 的回归；
断言（AssertionError: …）才指向真实的行为改变。两者的 FAIL 行长得一模一样，只有这一行能区分。

⚠️ 断言这一侧有一类例外，判据是断言在测什么，不是它是不是 AssertionError。 断言的对象是产品行为（一个值、一个形状、一次拒收）⇒ 照上面读：真实的行为改变，去查，⛔ 不要重排掉；
断言的对象是这次实验自身的有效性前提（跑完的耗时、负载下的先后、任何只在时间预算内才成立的条件）⇒ 它跟超时是同一类，同样对负载敏感，重排一次是合法的判别手段。
识别是机械的：断言的消息或它比较的值本身点名了一段时长、一个时间戳、一个耗时计数。实测过的一对 —— AssertionError: SecurityPlugin.init() ran: expected false to be true 测的是产品行为（真回归）；
AssertionError: this run took over a second, so second-precision stamps could have differed too: expected 1006 to be less than 1000 测的是实验前提：它守护的那条不变式当时是绿的，同一个 head 原样重排一次即成功。
穿着 AssertionError 外衣的时间测量，仍然是时间测量。（⛔ 这只改「怎么读一次红」，不改「哪些测试可以重排」——后者由别处管。）

跨 PR 相同签名（24h，按失败测试文件聚合）：

• src/apps/translations/setup-nav-dead-key-tombstone.test.ts — 24h 内已弹出 3 个 PR（i18n(platform-objects): ship the platform objects, Setup, Studio and Account in French #31233、feat(spec): declare the optional ISecurityService.describeEffectiveAccess and its EffectiveAccess answer types #31256、fix(lint)!: refuse a flow node no edge path from the start node reaches #31287），但它们同属 1 条投机栈 ⇒ 1 次独立命中。汇总 issue：Queue-flake anchor: src/apps/translations/setup-nav-dead-key-tombstone.test.ts #31306（已刷新）
• ⚠️ 24h 评论账本没读完（超过 5 页仍未读到窗口尽头），所以上面的「不同 PR 数」是下界，不是全量。

历史信号：

• ⚠️ 本 PR 过去 24h 已在队列失败 2 次（不含本次）。 内容未变而反复失败 ⇒ 高度怀疑 flaky 测试或与同组 PR 的语义冲突，重排不解决。
• 过去 24h 队列共有 25 个失败构建（不含本次）。

分诊清单：

1. 失败测试在本 PR 改动的包里 → 真回归，修 PR。
2. 失败测试与本 PR 无关 → 看上面的「跨 PR 相同签名」；已有汇总 issue ⇒ flaky/环境问题实锤，去那张 issue 上谈，修好前重排只会再烧一轮全队列。
3. 两者都不是 → 可能与同组 PR 语义冲突；等前面的 PR 落地或失败出队后再重排一次即可，不要连续重排。

---

Generated by Claude Code · merge-queue-triage workflow (#4859)

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 42430894350 红了。队列跑的是全量套件（PR 侧 CI 只跑 affected 子集），
所以失败的测试可能在本 PR 没碰过的包里 —— 那不是重排能修的。每次盲目重排都会让排在后面的所有 PR 重建一轮。

分类：failure —— 按下面的日志分诊。

失败的 job（日志抽取，best effort）：

• Test Core (3/8) — 失败步骤: Run this shard's tests

  @objectstack/platform-objects:test:  FAIL  src/apps/translations/setup-nav-dead-key-tombstone.test.ts > conditionally-contributed Setup nav ids stay labelled (#6659) > fr carries a label for every con
    ↳ 失败原因: @objectstack/platform-objects:test: AssertionError: Setup nav ids with no label — no boot-time gate can see these; see this file header: expected [ 'nav_scim_credentials' ] to deeply equal []
  

• Test Core (8/8) — 失败步骤: Run this shard's tests

  @objectstack/example-todo:test:  FAIL  test/loop-body-iteration.test.ts > #19206 — the loop body runs and binds `currentTask` > ESCALATION: the run reaches `notify_owner` and the notice renders the re
    ↳ 失败原因: @objectstack/example-todo:test: Error: Test timed out in 5000ms.
  @objectstack/example-todo:test:  FAIL  test/loop-body-iteration.test.ts > #19206 — the loop body runs and binds `currentTask` > EVERY item is processed, and each one carries its own values
    ↳ 失败原因: @objectstack/example-todo:test: Error: Test timed out in 5000ms.
  

↳ 失败原因 是判读的关键：超时（Test timed out in … / Hook timed out in …）多半是负载/时序，不是本 PR 的回归；
断言（AssertionError: …）才指向真实的行为改变。两者的 FAIL 行长得一模一样，只有这一行能区分。

⚠️ 断言这一侧有一类例外，判据是断言在测什么，不是它是不是 AssertionError。 断言的对象是产品行为（一个值、一个形状、一次拒收）⇒ 照上面读：真实的行为改变，去查，⛔ 不要重排掉；
断言的对象是这次实验自身的有效性前提（跑完的耗时、负载下的先后、任何只在时间预算内才成立的条件）⇒ 它跟超时是同一类，同样对负载敏感，重排一次是合法的判别手段。
识别是机械的：断言的消息或它比较的值本身点名了一段时长、一个时间戳、一个耗时计数。实测过的一对 —— AssertionError: SecurityPlugin.init() ran: expected false to be true 测的是产品行为（真回归）；
AssertionError: this run took over a second, so second-precision stamps could have differed too: expected 1006 to be less than 1000 测的是实验前提：它守护的那条不变式当时是绿的，同一个 head 原样重排一次即成功。
穿着 AssertionError 外衣的时间测量，仍然是时间测量。（⛔ 这只改「怎么读一次红」，不改「哪些测试可以重排」——后者由别处管。）

跨 PR 相同签名（24h，按失败测试文件聚合）：

• test/loop-body-iteration.test.ts — 24h 内已弹出 8 个 PR（feat(objectql): restore a recycle-bin entry under its original ids and emit data.record.restored (#30237 PR 3) #31220、i18n(platform-objects): ship the platform objects, Setup, Studio and Account in French #31233、feat(cloud-connection)!: refuse a managed upgrade that renames or retypes a stored field until it is declared or confirmed #31243、fix(auth): record an MCP access token as used only after the MCP door admits the call #31247、feat(plugin-auth, plugin-audit): a sign-in the IP allow-list refuses writes one login_failed row, bounded per address #31249、feat(spec): declare the optional ISecurityService.describeEffectiveAccess and its EffectiveAccess answer types #31256、feat(plugin-auth)!: an organization administrator reads their judged client address; the console's retired MFA notices are gone #31258、fix(metadata-protocol): a revert or a publish writes a connector's live row without the keys a save drains #31293），分属 4 条投机栈 ⇒ 4 次独立命中。汇总 issue：Queue-flake anchor: test/loop-body-iteration.test.ts #31294（已刷新）
• src/apps/translations/setup-nav-dead-key-tombstone.test.ts — 24h 内已弹出 4 个 PR（i18n(platform-objects): ship the platform objects, Setup, Studio and Account in French #31233、feat(cloud-connection)!: refuse a managed upgrade that renames or retypes a stored field until it is declared or confirmed #31243、feat(spec): declare the optional ISecurityService.describeEffectiveAccess and its EffectiveAccess answer types #31256、fix(lint)!: refuse a flow node no edge path from the start node reaches #31287），但它们同属 1 条投机栈 ⇒ 1 次独立命中。汇总 issue：Queue-flake anchor: src/apps/translations/setup-nav-dead-key-tombstone.test.ts #31306（已刷新）
• ⚠️ 24h 评论账本没读完（超过 5 页仍未读到窗口尽头），所以上面的「不同 PR 数」是下界，不是全量。

历史信号：

• ⚠️ 本 PR 过去 24h 已在队列失败 3 次（不含本次）。 内容未变而反复失败 ⇒ 高度怀疑 flaky 测试或与同组 PR 的语义冲突，重排不解决。
• 过去 24h 队列共有 31 个失败构建（不含本次）。

分诊清单：

1. 失败测试在本 PR 改动的包里 → 真回归，修 PR。
2. 失败测试与本 PR 无关 → 看上面的「跨 PR 相同签名」；已有汇总 issue ⇒ flaky/环境问题实锤，去那张 issue 上谈，修好前重排只会再烧一轮全队列。
3. 两者都不是 → 可能与同组 PR 语义冲突；等前面的 PR 落地或失败出队后再重排一次即可，不要连续重排。

---

Generated by Claude Code · merge-queue-triage workflow (#4859)

[objectstack-fleet]

domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx

Contract review — ACCEPT (delta)

Head-sha: e707147883c0a4f5b186eae38b7bba0e21a1a89c

ACCEPT · delta over dd9303ac18 (ACCEPT stands) · 2026-11-08T01:07Z

• f76d49c2 is a plain merge of origin/main. --remerge-diff shows no conflict resolution.
• e707147883 adds 5 lines in two files. In fr.ts, nav_scim_credentials: { label: 'Identifiants SCIM' } sits beside nav_sso_providers with the same conditional-contribution comment (plugin-auth: no Setup navigation entry reaches SCIM connection credentials, so the console's mint and revoke actions are found only by URL #31123). It is the short form of this locale's sys_scim_connection_credential.pluralLabel. fr.source-hashes.ts gets 2d94a32c33ca2957, the same en digest that ar, es-ES and ja-JP carry on main. This closes the deterministic queue red (anchor Queue-flake anchor: src/apps/translations/setup-nav-dead-key-tombstone.test.ts #31306) that ejected this PR three times and took stacked PRs with it.
• Dev report 6715403004: tombstone 15/15, platform-objects 1746 tests, check:app-nav-i18n, hub-parity 15, platform-page-i18n-parity 27, check:i18n, stale-fill and typecheck all exit 0. The cli plan.telemetry-sibling case passed 2/2 locally, so the shard 4/8 red in 42426481440 does not reproduce. The two GH006-refused push attempts changed nothing on the remote.
• os-zhuang's approval (on dd9303ac18) is not dismissed by a push on this ruleset. The seat re-enables auto-merge once CI is green on this head.