[objectstack-fleet]

Requested by Tesla · project thread

Part of #25712 (4b PR 2c: main's red Test UI 8/16 and 9/16 tests read what the tree has).
Part of objectstack#25677.

Clause-②: no
Metadata-protocol: none

维护者速读(草稿)

• 改了什么: objectui 那份 CI 说明页(ui/content/docs/guide/ci-cd-pipeline.md)删掉已被 release(ui-merge): one release lane for both scopes — @object-ui joins the fixed group, @objectstack/console builds from ui/apps/console, the objectui pin is deleted (#25711) #25745 删除的两个 workflow(changeset-release.yml、spec-range-floors.yml)的清单行、章节和 NPM_TOKEN 行;两个测试文件不再去读这两个已不存在的文件。
• 为什么改: main 上 Test UI 8/16 和 9/16 因此是红的(11 个失败)。本 PR 后只剩 1 个:十个嵌套 workflow 还在调用已删除的 scripts/objectstack-pin.mjs。这是真问题,由 2d 删除 ui/.github/workflows/** 时一并清掉。
• 风险与代价(含回滚): 只动一页文档和两个测试,不动任何会运行的代码或受管面;回滚 = revert 本 PR。
• 席位意见:(留空)
• 你要做的: 无,Tier S,由席位落地。

席位核对

(left for the dispatch seat)

Why

objectstack#25745 (4a) deleted ui/.github/workflows/changeset-release.yml and ui/.github/workflows/spec-range-floors.yml when ui/ took objectstack's one release lane for both scopes. scripts/ui-import/README.md recorded that three ui/scripts/__tests__ files still read them and left them to this card. On main (90a34e20, e167956e), and reproduced locally:

• Test UI 8/16: ci-cd-pipeline-doc.test.ts, 9 failures (ENOENT … ui/.github/workflows/changeset-release.yml; inventory rows for both deleted files).
• Test UI 9/16: workflow-cache-save-bound.test.ts, 2 failures (changeset-release.yml must still define a release: / lane: job).

The two workflows stay deleted. Each fix makes the page and the tests describe the tree as it is. Nothing is skipped: rows whose whole subject is gone retire, and the disposition is written beside them.

What changes

• ui/content/docs/guide/ci-cd-pipeline.md (+16 / −240):
  • The two inventory rows, the ### Changeset Release section and its four sub-sections, the ### Spec Range Floors section, and the NPM_TOKEN row (its only user was the deleted workflow) are removed.
  • Historical prose that named changeset-release.yml now says "the Changeset Release workflow". The page's own test reads every bare name.yml as a claim that the file exists.
  • The two present-tense lists that included spec-range-floors.yml drop it.
  • ## Release Workflows opens with one paragraph saying the two workflows were deleted by release(ui-merge): one release lane for both scopes — @object-ui joins the fixed group, @objectstack/console builds from ui/apps/console, the objectui pin is deleted (#25711) #25745 and that @object-ui/* now publishes in objectstack's release lane.
• ui/scripts/__tests__/ci-cd-pipeline-doc.test.ts (+16 / −57). In "the four sections measured as parity defects":
  • the two units (changeset-release.yml release and spec-range-floors.yml) and their entries in the jobs-per-workflow pin come out;
  • so does the lane exclusion check;
  • so do the four DECLARED_NON_RUN_COMMANDS entries that explained the release job's publish: input.
  • The sweep's inherited-declaration pin goes from five keys to one.
  • A paragraph beside the set records the retirement. The two parser floors (> 8) still hold on what is left.
• ui/scripts/__tests__/workflow-cache-save-bound.test.ts (+11 / −2): changeset-release.yml::release (40) and changeset-release.yml::lane (20) leave their two derived-ceiling tables. Each removal is recorded the way the file already records stale.yml::stale's: "re-adding it does not restore a pin".

The one red that stays, owned by PR 2d

Test UI 8/16 keeps one failure after this PR: ui/scripts/__tests__/check-pre-install-import-graph.test.ts > finds no pre-install gate reaching a package.

AssertionError: a script this repository runs before `pnpm install` needs `node_modules` to load: expected [ Array(1) ] to deeply equal []
+   "unresolved: scripts/objectstack-pin.mjs",

That is a true finding on the tree, not a stale pin. #25745 deleted ui/scripts/objectstack-pin.mjs, but ten nested workflows still run node scripts/objectstack-pin.mjs build before their install step: ci.yml, docs-deploy.yml, lint.yml, live-e2e.yml, lockfile-dedupe.yml, node-esm-load-gate.yml, performance-budget.yml, published-dist-gate.yml, repo-gates.yml and shadcn-check.yml under ui/.github/workflows/. Those files are Tier H since #25834, never run nested, and are deleted by PR 2d. Per the dispatcher's ruling, there are no throwaway edits to them here. The finding is the first row of 2d on #25712.

Verification

Run from ui/ per ui/AGENTS.md:

Run	This PR	main e167956e
The three files: pnpm exec vitest run --project unit scripts/__tests__/ci-cd-pipeline-doc.test.ts scripts/__tests__/workflow-cache-save-bound.test.ts scripts/__tests__/check-pre-install-import-graph.test.ts	1 failed / 132 passed (the owned row above)	12 failed / 121 passed
ci-cd-pipeline-doc.test.ts alone	85 passed	9 failed
The whole scripts/__tests__ suite, --shard=1/2 and --shard=2/2	26 failed / 5,576 passed	—

Of the 26 suite failures, one is the owned row. The other 25 are date-only-zone-pins-10183 ×23 and unit-registry-absence-collision ×2, and they fail identically on a clean main e167956e checkout here (25 failed / 19 passed for those two files). Their spawned runs report "ran no case at all" on this unbuilt box, so they are not this diff's.

Derived gates (node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack from this diff): 18 families, all exit 0 at this head; --ran reconciles 18 run, 0 NOT-MEASURED, 0 UNRUN. The slowest were check:query-options-erasure (285 s), check-comment-mask-corpus (206 s) and check:type-check-debt (185 s).

🤖 Generated with Claude Code

https://claude.ai/code/session_01A6SEitNngAMugGbuJHWsDS

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 40255506383 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 UI (8/16) — 失败步骤: Run the ui tests (shard 8/16)

  × finds no pre-install gate reaching a package 12ms
  FAIL   unit  scripts/__tests__/check-pre-install-import-graph.test.ts > the real tree — every step objectui#6148 measured is still derived > finds no pre-install gate reaching a package
    ↳ 失败原因: AssertionError: a script this repository runs before `pnpm install` needs `node_modules` to load: expected [ Array(1) ] 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）：

• scripts/__tests__/check-pre-install-import-graph.test.ts — 24h 内已弹出 5 个 PR（fix(ui-merge): two tests that #25745 broke on main read the merged tree — the Changesets config through changesetConfigPath(), the SDUI html tier by its tags (#25711) #25852、ci(ui-merge): Type Check UI builds the five vitest-setup packages before that step, so an empty ui scope no longer reds it #25856、ci(ui-merge): register Lint UI and Build & E2E UI in REQUIRED_CONTEXTS #25857、fix(ui-merge): the CI page and two of its tests stop reading the two workflows #25745 deleted #25861、test(ui): three ui/ script tests spell their temp bases so the whole-tree scratch-dir sweep reads them (PM dispatch-gates red on main) #25865），但它们同属 1 条投机栈 ⇒ 1 次独立命中。汇总 issue：Queue-flake anchor: scripts/__tests__/check-pre-install-import-graph.test.ts #25863（已刷新）

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 7 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)