[objectstack-fleet]

Requested by Bill · project thread

domain:server#6 · os-bill · session_01Utp6CXCS98uhsFQYk64epE

Fixes #30164

Clause-②: no
Metadata-protocol: none
Skills-docs: none (git grep -n 仪表板 -- skills/ on the final tree: 0 hits)

维护者速读

• 改了什么：平台下发的中文仪表盘编辑表单里 6 处“仪表板”改为“仪表盘”（表单标题、基础信息说明、页眉及其显示标题/显示描述、组件的帮助文字）。
• 为什么改：控制台和 Studio 已统一叫“仪表盘”，只有平台下发的这张表单还叫“仪表板”，同一对象一个流程里两个名字。
• 风险与回滚：只改中文显示文字，不动任何键、结构或其他语言；回滚即还原这一个文件。
• 先看：packages/platform-objects/src/apps/translations/zh-CN.metadata-forms.generated.ts（6 行文字），再看新测试 zh-cn-dashboard-word.test.ts。
• 两分钟验证：pnpm --filter @objectstack/platform-objects exec vitest run src/metadata-translations/zh-cn-dashboard-word.test.ts，预期 2 条通过。
• 需要你做的：无，席位落地。

What changes

zh-CN.metadata-forms.generated.ts is the os i18n extract merge baseline (scripts/i18n-extract.config.ts pins metadataForms to the committed file, and --merge only fills gaps), and both its own header and metadata-translations/index.ts say its leaves are edited in place. So the six dashboard leaves were corrected there, then the bundle regenerated: pnpm i18n:extract reproduces the committed tree byte for byte (no diff after the run) and pnpm check:i18n is green. No other source under packages/ carries the zh-CN form text. The source-hash tables are untouched: these leaves are translations, not source copies, so the generated provenance table never recorded them.

Tests

New packages/platform-objects/src/metadata-translations/zh-cn-dashboard-word.test.ts reads the SERVED bundle (MetadataFormsTranslations['zh-CN'], after withSourceFallback, the bundle rest-server.ts localizes /api/v1/meta/types forms with): the dashboard form title is 仪表盘, and no served zh-CN metadata-form string contains 仪表板.

• Red on main's bundle: both tests fail, the second listing the six strings.
• Green after the fix; ablation (one leaf put back via scripts/ablation-replace.mjs): the class test goes red naming 仪表板标识与描述, restored to the HEAD blob with git diff HEAD empty.
• pnpm --filter @objectstack/platform-objects test: 112 files, 1523 tests passed. typecheck (includes check:test-typecheck): exit 0.

Acceptance notes

• packages/rest/src/meta-types-schema-titles.test.ts and packages/runtime/src/domains/i18n-translations-served-groups.test.ts name 仪表板 only inside their own inline fixture bundles and assert the fixture back; they do not read the platform bundle, so they were left as they are.
• packages/spec/src/marketplace/package-l10n.test.ts and packages/spec/src/system/i18n-resolver.test.ts likewise use their own fixtures and do not pin served form text (out of this card's surface, untouched).

Verification

All local readings on head df4de5fb (after merging origin/main).

• pnpm check:i18n: exit 0, check-i18n-bundles: OK (10 package(s) — all bundles in sync, no undeclared authoring keys).
• pnpm i18n:extract: exit 0, 11 files written, git status clean afterwards.
• pnpm --filter @objectstack/platform-objects test: exit 0 (112 files, 1523 tests). typecheck: exit 0.
• pnpm lint: exit 0.
• dispatch-gates --commands: 58 derived; --ran reconciliation: 58 accounted, 56 exit 0, 2 NOT MEASURED (exit 3, PREREQUISITE NOT MET):
  • node scripts/check-plugin-teardown-shape.mjs --self-test: shallow clone cannot reach its pinned positive-control commit (checker-health only, judges no diff).
  • pnpm check:dual-build-cjs-loads: needs every workspace package's dist/; only the i18n build closure was built here. CI runs both.

Red on main, ablation

Red (main's bundle): expected '仪表板' to be '仪表盘'; offenders listed 仪表板, 仪表板标识与描述, 仪表板页眉配置（title、subtitle、actions）, 在页眉中显示仪表板标题, 在页眉中显示仪表板描述, 包含位置和尺寸的仪表板组件. Tests 2 failed (2).

Ablation: ablation-replace: ok mutation landed: anchor 1 -> 0, then × no served zh-CN metadata-form string carries the retired spelling 仪表板, Tests 1 failed | 1 passed (2); ablation-replace: ok restored: blob == HEAD (ab03233a6fe0) and git diff HEAD is empty.

---

Generated by Claude Code

🤖 Generated with Claude Code

https://claude.ai/code/session_01Utp6CZNuVoLVGgYU36KdYe

[github-actions]

📓 Docs Drift Check

Nothing in this diff resolved to a documentable surface (no symbol, route or SDK anchor derived from 1 changed package(s)), so this run has no opinion about the docs.

What this run could not see

• 1 name(s) were too generic to anchor anything (single lowercase words)
• 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 — 2 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 a1d0b1ee9e428402399f91808923c508c84e0957 → packageMentionDocs.

[objectstack-fleet]

domain:server#6 · os-bill · session_01Utp6CXCS98uhsFQYk64epE

席位核对 · ACCEPT · 2026-11-05T13:18Z · head df4de5fb

Reviewed against order server-seat6/dispatch-30164.md (Claim 6658014415), against the PR's own file list and diff.

• Shape. Six zh-CN leaves in zh-CN.metadata-forms.generated.ts change 仪表板 → 仪表盘 (form title, basics description, header, show-title, show-description, widgets help); no key, structure or other locale moves. The file is the extract's merge baseline and its header says to edit leaves in place, the route the seat confirmed in the Claim; pnpm i18n:extract reproduces it byte for byte and pnpm check:i18n is green (dev readings).
• Pins. zh-cn-dashboard-word.test.ts reads the served bundle (MetadataFormsTranslations['zh-CN'] after source fallback): red on main, green here, one ablation recorded.
• Declarations. Clause-②: no, Metadata-protocol: none, patch, Skills-docs: none (grep 0). Not BREAKING.
• Accepted deviations. none. Two derived gates read NOT MEASURED locally (shallow-clone / full-dist prerequisites); CI gates the queue.
• Out of scope. Fixture-only mentions of 仪表板 in packages/rest, packages/runtime and packages/spec tests do not pin served text: Acceptance notes, nothing to file.
• Gates. check-governed-merges --pr 30485: not governed, 54 changed lines. Author objectstack-fleet[bot] (relay).

Landing: ready + auto-merge; the queue waits for every check.

[objectstack-fleet]

domain:server#6 · os-bill · session_01Utp6CXCS98uhsFQYk64epE

席位记录 · 2026-11-05T13:49Z · Head-sha df4de5fb · 补记(ACCEPT 6659011574 之后)

维护者速读(席位按 diff 校正)

• 改了什么：平台下发的中文仪表盘编辑表单里 6 处“仪表板”改为“仪表盘”（表单标题、基础信息说明、页眉、显示标题、显示描述、组件的帮助文字）。
• 为什么改：控制台和 Studio 已统一叫“仪表盘”，同一对象在一个流程里不该有两个名字（Studio on zh-CN still mixes languages and terms: view properties panel, flow nodes, publish sheet labels, and 仪表板 beside 仪表盘 #30039 第 4 项）。
• 风险与回滚：只改中文显示文字，不动键、结构和其他语言；回滚即还原这一个文件。
• 先看：zh-CN.metadata-forms.generated.ts 的 6 行，和新测试 zh-cn-dashboard-word.test.ts。
• 两分钟验证：pnpm --filter @objectstack/platform-objects exec vitest run src/metadata-translations/zh-cn-dashboard-word.test.ts，2 条通过。
• 需要你做的：无。非受管面、非 BREAKING，技能不要求你审，席位落地。

席位意见：同意落地。按文件头规定的做法原地改基线叶子再重生成，重生成逐字节一致，测试读的是下发后的包。

席位核对

1. 形态：draft→ready，目标 main，Fixes #30164；作者 objectstack-fleet[bot]（中继）；正文无闭合关键词挨着别的卡号。
2. 前提实测：origin/main c5800818 上该文件 6 处 仪表板（席位 grep）；PR diff 正好这 6 行。
3. 门禁 check-runs(钉 head df4de5fb)：49 条，36 success、13 skipped、0 失败；Lint & Repo Gates success、TypeScript Type Check success。
4. 接受的偏差：无。
5. 范围外去向：rest/runtime/spec 测试里的 仪表板 是自带 fixture，不读下发文字，Acceptance notes，不立卡。Skills-docs: none（grep 0）。changeset patch，Clause-②: no，Metadata-protocol: none。
6. 落地条款：check-governed-merges --pr 30485：NOT governed，54 行。已 ready + auto-merge，13:43Z 入合并队列（排在 main-red 止血 PR fix(cloud-connection): stage the ledger adoption copy without mkdtempSync, so main's dispatch-gates lane is green again #30487 之后）。合并后席位写落地、摘标。

[objectstack-fleet]

domain:server#6 · os-bill · session_01Utp6CXCS98uhsFQYk64epE

出队记录 · 2026-11-05T14:09Z · head df4de5fb

Dequeued by the merge queue (CI_FAILURE). Signature, read from the group's job log (run 42040560124, job 127290410036):

• Live E2E UI › e2e/live/studio-customize-journey.spec.ts:121 › "a read-only base is customized, restored and extended through Studio, and its uninstall is refused" → apiRequestContext.get: read ECONNRESET on GET /api/v1/meta/view/showcase_task.in_progress (16 of 17 live specs passed).
• Not this PR's: the diff is six zh-CN translation leaves and one test in packages/platform-objects; the failing spec reads an en view through /meta/view. The same signature failed the group of PR refactor(objectql): type the six #4918 erasure-pinned ObjectQL methods and move their bodies out of engine.ts with the codemod's fourth pass #30463 on the same base fa9c0aa2 (run 42037558126, job 127279783258), so it reproduces on the queue base without this diff.
• Initial judgement: a new signature on main fa9c0aa2, not a known flake in the ledger. ⛔ Not requeued blind: Request: urgent-triage sent to triage#1's mailbox. The seat requeues once the base is green again or triage rules it a flake.

[objectstack-fleet]

domain:server#6 · os-bill · session_01Utp6CXCS98uhsFQYk64epE

重投 · 2026-11-05T14:17Z · head df4de5fb：分诊 triage#1 判该签名非确定性（叠加组绿），已立 #30524（p1，domain:devx）承接 studio-customize-journey ECONNRESET（信箱 6660234942）。按已知 flaky 原样重投一次：重新挂 auto-merge。同签名再弹即停，交 #30524。

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 42040560124 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）：

• Live E2E UI — 失败步骤: Run the live e2e specs（日志不可读，点进 job 看）

↳ 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）：

• ⚠️ 本次没有可用的聚合签名（日志里没有能解析出测试文件名的 FAIL 行）—— 这不是「没有同签名的其他 PR」，是这一轮没测到。跨 PR 聚合本次不可用，请手工比对其他 PR 的同类评论。
• ⚠️ 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 15 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)