chore(pm): unstale the release-aftercare ceiling comment in the line ratchet

[os-litant]

Fixes#13093

Comment-only. The gate's predicate, every ceiling value and the CEILINGS data are unchanged; the gate was green before and stays green.

What was stale

The comment block attached to the release-aftercare.mdCEILINGS entry carried two claims that were true when written and went stale when landing-operations.md was re-pinned 82 to 80.

Before:

// count (headroom 0, same convention as the entries above). Its pointer from// landing-operations.md rides existing slack on that file's last// MERGED-tracking line, so that ceiling stays at 82 — no re-wrap, no cut.

After:

// count (headroom 0, same convention as the entries above). It is pointed at// from landing-operations.md, whose ceiling is the entry directly above and is// deliberately not restated here: a live number copied into a neighbour's// comment goes stale the next time that file moves — which is what happened to// the arithmetic this sentence replaces.

Evidence that both claims were stale

Claim 1 — "that ceiling stays at 82". The gate prints the refutation on every run:

✓ check-skill-line-ratchet: .claude/skills/pm-dispatch/references/landing-operations.md is 80 lines (ceiling 80; headroom 0).

The CEILINGS entry immediately above the comment reads 80, and this file's own header already records the move — "landing-operations.md moved the other way (82 → 80, its standing headroom locked in)". The comment therefore contradicted both its neighbour and its own header.

Claim 2 — "rides existing slack on that file's last MERGED-tracking line". Measured against the script's own MAX_LINE_BYTES (120): the pointer to release-aftercare.md sits on line 49 of landing-operations.md at 104 bytes, while the last MERGED-tracking line in that file is now line 53. The positional claim no longer locates the pointer, and the file sits at its ceiling with headroom 0.

Why the arithmetic was dropped rather than corrected

Restating a neighbouring entry's live number is precisely what made this comment stale. A corrected number would re-stale the next time landing-operations.md moves, so the smaller honest wording wins: the CEILINGS entry and the gate's per-run verdict line are both authoritative and self-updating, and the comment now points at them instead of copying them. The replacement sentence also records why the arithmetic is absent, so a later author does not helpfully re-add it.

Gates

Union derived mechanically with node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack; all runs below are at commit 08caf7243, which is this branch's head. Exit codes captured by redirect-then-capture, never through a pipe.

Gate	Trigger	Result
pnpm check:pm-skill-ratchet (includes --self-test)	gate script identity	green — ✓ check-skill-line-ratchet self-test: 111 cases pass.
pnpm check:agent-test-spelling	scripts/**	green
pnpm check:bash32-floor	scripts/**	green
pnpm check:cli-command-ids	scripts/**	green
pnpm check:cross-package-test-inputs	scripts/**	green
pnpm check:entry-guard	scripts/**	green
pnpm check:parse-guard	scripts/**	green
pnpm check:pnpm-filter-targets	scripts/**	green
pnpm check:watch-hint-literal	scripts/**	green
node scripts/check-ci-filter-parity.mjs	scripts/**	green
node scripts/check-cross-package-test-inputs.mjs	scripts/**	green
node scripts/check-shard-attestation.mjs	scripts/**	green
node scripts/pm/bare-root-worklist.mjs --self-test	convention: edits a gate script	green
pnpm check:pm-dispatch-gates	convention: edits a gate script	green
node scripts/check-published-list-mirrors.mjs	reads the edited file	green
node scripts/check-skills-token-ratchet.mjs	reads the edited file	green
node scripts/check-test-completeness.mjs	scripts/**	not measured locally — CI-only

check-test-completeness.mjs has no pnpm wrapper and takes a turbo test log produced at runtime by the Test Core job (node scripts/check-test-completeness.mjs "$RUNNER_TEMP/test-core.log"). Invoked bare it prints its usage line and exits 1, which is an invocation error rather than a verdict, so it is recorded as not measured rather than as a pass or a failure. CI runs it with its real argument.

The last two rows are not path-derived: they are the scripts that textually reference the edited file, run because a comment change inside a gate script is exactly the kind of edit a mirror or token gate could read.

Changeset

None. scripts/pm/** is a repo-internal gate surface that publishes nothing, and this diff is comment-only, so the PR takes the skip-changeset label instead.

Scope

One file, one hunk, five lines replacing three. No ceiling value, predicate or CEILINGS datum is touched — including the SKILL.md row-pin that a later queued card will lower.

Generated by Claude Code

---

Generated by Claude Code

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 33242316964 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。

失败的 job（日志抽取，best effort）：

• Test Core (1/6) — 失败步骤: Run this shard's tests

  @objectstack/cli:test: FAIL test/serve-publishes-bound-port.e2e.test.ts > #13062 `os serve --port 0` — the request that can never be the answer > announces the BOUND port on all three channels, and ↳ 失败原因: @objectstack/cli:test: Error: ENOENT: no such file or directory, open '/tmp/os-bound-port-home-jzl4eA/runtime.env_local.json'
  

↳ 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。

跨 PR 相同签名（24h，按失败测试文件聚合）：

• test/serve-publishes-bound-port.e2e.test.ts — 24h 内已弹出 10 个 PR（Hold the isSystem census page to the code: committed AST census, population gate, converted anchors #13051、fix(ci): count speculative-stack inheritance as one hit in merge-queue triage #13107、fix(rest): the /data declared-4xx body carries human language in error, not the ADR-0111 CODE: prefix #13124、feat(metadata): the artifact door registers stack-declared capabilities (#12892 step 1) #13125、fix(tooling): declare kebab-case diagnostic codes a separate vocabulary, in one place #13130、docs(kernel): teach the dot-qualified view item name in metadata-service.mdx #13133、docs(devx): correct the stale collector claim on the js-comment-mask entry #13140、fix(scripts): close an ADR-0087 framed region at the same or a shallower heading, not at any heading #13142、feat(pm): record a passive boot marker on every verify-lock invocation, so cross-restart evidence about /tmp accumulates by itself #13145、chore(pm): unstale the release-aftercare ceiling comment in the line ratchet #13161），分属 8 条投机栈 ⇒ 8 次独立命中。汇总 issue：Queue-flake anchor: test/serve-publishes-bound-port.e2e.test.ts #13158（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 无队列失败记录（首次）。
• 过去 24h 队列共有 10 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)

[os-litant]

队列失败处置(skills 席):非本 PR 之失,不盲目重排。本 PR 是 scripts/pm/check-skill-line-ratchet.mjs 纯注释 diff,不可能触及 @objectstack/cli 的 e2e;失败签名(serve-publishes-bound-port.e2e.test.ts,ENOENT 于临时目录,非断言)24h 内已命中 10 个 PR / 8 条独立栈,属 #13158 汇总的既知环境 flake。按分诊单第 3 条:等该 flake 修复或本波队列消化后重排一次(本席 08:48Z 检查点执行,或维护者顺手)。不在本 PR 上修 cli 测试——那是 #13158 的。

---

Generated by Claude Code