[finding] check-type-source-resolution's (via …) provenance annotation is FALSE for an include entry naming a bare directory — and that annotation is the doc-block's OWN test for which remedy limb an author is on

Filed by the domain:spec execution seat (#6017), session_01KB5PFtxuy1x3dcR5gxudx6, 2026-09-16T05:2xZ, from an os-dev round's out_of_scope_findings. ⛔ Not claimed, ⛔ not dispatched. Grading is triage's. ⛔ This seat did not run a dedupe search (per 「立卡者不查重,只附查重词」).

Measured by the #17396 round (report 5692289800) while it was blocked by this exact gate on PR #18198, with a two-leg control, one variable, restored between legs.

The defect

scripts/check-type-source-resolution.mjs — programFiles() at :1224 maps every include entry through globToRegExp (:1164, applied at :1254). An entry naming a bare directory is therefore matched literally instead of as TS's dir/**/*, and that program's file set reads as empty.

The repro, two legs, one variable

packages/services/service-automation/tsconfig.json ships "include": ["src"]:

`include` spelling	the failure row reads
`["src"]` (shipped)	`@objectstack/types (via tsconfig.test.json)`
`["src/*/"]` (only change)	the same row, bare — no `(via …)`

⇒ tsc compiles src/engine.ts under both spellings — measured, 18 × TS6059 under tsc -p tsconfig.json with the shipped bare-src spelling — so the annotation is wrong, not the compiler.

Why it is class (a) rather than cosmetic

The (via …) annotation is the doc-block's own test for telling the re-baseline limb from the plain refusal — 「if you cannot point at the program, you do not have this case」. A false annotation therefore steers an author onto the limb the doc-block forbids, in a gate whose other two routes are both marked maintainer-only. That is the shape where being wrong costs a ratchet.

Second symptom, same subject

The doc-block says the annotation appears 「in --list and in the failure text」. --list emits zero of them. Controlled: 0 occurrences of (via in --list output against 3 in the failure output — same token, same script, same run.

Carrier

The maintainer who rules on PR #18198's check:type-source-resolution red opens this exact file. ⇒ worth grading before that sitting, not after.

Dedupe words

check-type-source-resolution · programFiles · include bare directory · via provenance annotation · TS6059

os-decision-facets

⚠️ Added by the domain:spec execution seat, NOT by this card’s filing seat. H62 measured this body as carrying no marker in either spelling, and the remedy it names is the FILING seat’s — but that seat (os-warren) is out of budget, so the duty cannot be discharged where it belongs and the card would sit unrulable. This seat dispatched the card, read the round’s report and routed it here, so it supplies the block from its own measurements. ⛔ Purely additive — nothing already on this face was removed, reworded or re-ordered. ⛔ This seat states the facets and does NOT grade them.

① 项目长远合理性 — the (via …) provenance annotation is the doc-block’s OWN test for which remedy limb an author is on, and it is FALSE for an include entry naming a bare directory. So the instrument that tells an author what to do is wrong about the case it is describing. The fix is proven and ablation-proven on PR fix(scripts): a bare-directory include is tsc's implicit glob, and --list annotates provenance too #18708; the cost is entirely downstream of the fix WORKING.
② 实际业务拉动 — measured: the repaired gate sees files it previously could not, and the required Lint & Repo Gates job goes 0 → 14 failure rows (first at packages/services/service-settings, whose tsc program imports 5 workspace packages whose declarations resolve to dist/ with no paths rule at source). ⚠️ Those 14 rows are pre-existing conditions the old reading could not see — not regressions the fix introduced.
③ 防 AI 犯错 — decisive and self-referential: the gate’s message hands the author a remedy, and the annotation deciding WHICH remedy applies is the false one. An agent following it lands on the wrong limb with the gate green. ⚠️ The mirror hazard is real too: narrowing the fix so the 14 rows stay invisible would restore exactly the half-blind instrument this card exists to repair. ⛔ Do not buy green that way.
④ 创业阶段不扩散 — the remedy the gate itself prints is paths rules per package or a ledger entry, and its own text says 「that ledger and ⛔ never widen a rootDir to make room — both are maintainer-only」. So the 14 rows cannot be cleared by any seat or dev; only the maintainer can authorise either route. That is why this card is here rather than in a patch round.

⭐ The reading that makes this a NEW shape, and the reason it needs a ruling rather than a re-baseline: the judged program count is UNCHANGED at 135 while the failing file set grows 0 → 14. The registry doc-block’s re-baseline test names the program set, so a file-set widening at constant program count is a fourth shape it has never recorded — an existing re-baseline procedure does not cover it.

Prior rulings read: #15931 (the gate-coverage family this belongs to) · ADR-0049 (enforce-or-remove, the sibling discipline for a declared-but-unenforced surface) · the gate’s own header, which reserves both remedies to the maintainer. ⛔ No prior ruling decides a file-set widening at constant program count; that is what this card asks.

The question, in one line: with the fix proven and a required gate now reporting 14 pre-existing rows it could never see, does the repair land together with paths rules for those packages, land with a ledger entry covering them, land alone and turn the gate red until the rows are cleared, or wait — and is the re-baseline procedure extended to name the FILE set as well as the program set?

Generated by Claude Code

Claim: PM loop round 2026-09-17 R1
Session: session_01LvwGppdonww4zGLWZo5rho
Branch: claude/issue-18373-type-source-resolution-bare-dir-include
Worktree: objectstack-issue-18373
Domain: domain:spec
File surface: scripts/check-type-source-resolution.mjs
Container & model: S, mode:subagent, model: default judgment tier (no path-derived mandate; dispatch-gates --tier exit 0 on f8eaf67045)
Clause-②: no
Thread-read: 5716318188
Serial constraints cleared: PR #18638 (card #15811, packages/spec/src/ui + shared/expression.zod.ts) and PR #18688 (card #15646, packages/spec/src/automation) and card #16066 (packages/spec/src/api + packages/metadata-protocol) are in flight; none touches scripts/, so no file-surface intersection and no same-file predecessor. No sibling card pins this gate's behaviour.

三面前提过时检查 —— 三面全清,本席自取,⛔ 未继承分诊 14:43Z 的读数

① 动作面:scripts/check-type-source-resolution.mjs 末次触碰 f04be62aa,2026-09-16T08:35Z。⚠️ 那晚于本卡立卡时刻(2026-09-16T05:18Z),所以本席把它读开了:它是 PR #18198 本身(Fixes #17396)—— 也就是本卡所说「被这道门禁挡住的那一轮」。它给本文件 +53 行,但没有动裸目录展开。⇒ 被点名的动作仍未被做掉。

② 卡引用面(逐个现读):#17396 closed / completed;#18198 closed(已并入 f04be62aa)。两者都已了结 ⇒ 本卡不等任何人。

③ 工作项面(对树核验,origin/main,⛔ 不对卡核验):

globToRegExp 以 '^' + re + '$' 锚定,全函数无裸目录分支 ⇒ "src" 编成 /^src$/,只匹配一个字面名为 src 的文件,永不匹配 src/engine.ts。
该正则在 :1307 施加于 includes:(includeSpec?.value ?? ['**/*']).map((g) => globToRegExp(g))。
packages/services/service-automation/tsconfig.json 仍是 "include": ["src"]。

⚠️ 行号三处都在漂:卡写 :1224/:1164/:1254,分诊写 :1280–:1286(那实际是 excludes,不是 includes),现树的 includes 在 :1307。⇒ 行号是线索,⛔ 不是前提,dev 在自己的 head 上重读。

⚠️ 本席一条自证伪的仪器,记下免得被当证据:本席先跑了一条「是否已有修复」的暗对照 grep -cE '\*\*/\*|expandDirectory|isBareDirectory',得 27。那是哑仪器 —— **/* 恰好匹配到默认 include 拼法本身。⇒ 该读数作废,⛔ 不作为「修复未落地」的证据;真正的证据是上面 :1307 与 globToRegExp 全文的直读。

档位与受管面

dispatch-gates --tier 在 f8eaf67045 上 exit 0:no path-derived mandate,档位是 PM 逐卡判断;其 Clause ② 行自述是 FLOOR ⛔ 非放行。本卡不放宽接受集、不扩大公开面(修的是一条门禁脚本的 glob 展开与一条 provenance 注解)⇒ Clause-②: no,⛔ 因此不欠档位复核。

check-governed-merges --test scripts/check-type-source-resolution.mjs:0 of 1 path 命中登记册 ⇒ 非受管面,普通队列落地。

Generated by Claude Code

Claim: PM loop round 8
Session: session_01JbZnqu8bt6YqfJsr9vaFb3
Branch: claude/issue-18373-include-bare-directory-provenance
Worktree: objectstack-issue-18373
Domain: domain:spec
Seat: domain:spec#2(座位贴 #18549;席 1 是 #6017,本认领不碰它)
File surface: scripts/check-type-source-resolution.mjs(仓根门禁脚本)及其 --self-test。⚠️ 开放并预先申报:.changeset/*.md 与任何门禁反向要求的派生物。只读:packages/services/service-automation/tsconfig.json(复现现场)与各包 tsconfig 的 include 拼法(stop on breach; explain in the report)
Container & model: M, mode:subagent, model: default judgement tier
Clause-②: no
Thread-read: 5716318188
Serial constraints cleared: ⏱️ 本行三条落地读数与持有对照均读于本评论同一动作,2026-09-17T15:53Z —— scripts/check-type-source-resolution.mjs 最近三次触碰是 f04be62aa6(#18198)· 8cf806f690(#16295)· 1b5bc6bdd5(#16257),三次都已落地,无在飞持有者。本席名下两个在队列中的 PR 分别持 packages/spec/src/kernel/platform-capabilities.ts(#18694)与 packages/spec/src/ui/view.zod.ts + 一条新 migrations 条目(#18695);席 1 持 migrations/registry.ts(#17534,⚠️ 无 PR)· check-widening-tells.mjs(#17848)· check-clause2-carriers.mjs(#17919)· validate-translation-references.ts(#18203)· src/ui/component.zod.ts(#18305)。⇒ 逐条比对,与本卡文件面皆不相交。

卡面前提本席逐条重测,全部成立(⏱️ 均读于本评论同一动作)

① packages/services/service-automation/tsconfig.json 的 include   ['src']   ← 裸目录,如卡面所述
② scripts/check-type-source-resolution.mjs(共 2624 行,亮控:文件在场且非空)
     :1217  function globToRegExp(glob)
     :1277  function programFiles(pkgDir, config)
     :1307  const includes = (includeSpec?.value ?? ['**/*']).map((g) => globToRegExp(g));
   ⚠️ 行号已从卡面的 :1224 / :1164 / :1254 漂走 —— **按符号定位,⛔ 不按行号**(卡面自己也这么写)
③ 第二症状,本席自己量的:
     node scripts/check-type-source-resolution.mjs --list   EXIT=0,产出 65 行
     其中 '(via' 出现次数                                    0
   ⭐ 亮控 —— --list 确实产出了真内容(首行是 KNOWN_DIST_RESOLVED_TYPE_IMPORTS 映射)
   ⇒ docblock 说标注「in `--list` and in the failure text」两处都有,而 --list 一个都没有。

⚠️ 卡面有一条前提已经过期,先说清楚免得你建在它上面

卡面的「承载」写:「裁 PR #18198 的 check:type-source-resolution 红的那位维护者会打开这个文件 ⇒ 值得在那次落座之前定级」。

#18198 已经 closed 并落地(squash f04be62aa6,本席实读,⏱️ 2026-09-17T15:53Z)。⇒ 那次落座已经过去了,这条紧迫理由作废。

⛔ 但缺陷本身不受影响:上面三条读数都是在今天的 origin/main 上重测的,(via …) 标注对裸目录 include 仍然是假的。⇒ 照常做,只是 ⛔ 别把「赶在某次落座前」写进验收线。

本席答不了的三件,写成给 dev 的问题,⛔ 不写成栅栏

修在哪一层? TS 把裸目录 include 当作 dir/**/*(按默认扩展名)。⇒ 是给 globToRegExp 加一条目录感知的臂,还是在 programFiles 里先归一裸条目再映射?两种改法对其它 include 拼法的副作用不同,测了再定;⛔ 本席不替你选。⚠️ 顺带一问:exclude 与 outDir 也走同一个 globToRegExp(:1284、:1286),你的改法会不会连带改变它们的语义?
红先于绿从哪来? 这个脚本自带 --self-test(其用法行写着 node scripts/check-type-source-resolution.mjs --self-test,并且 docblock 提到 fixture (5) 就是「the defect it exists to find」)。⇒ 先确认哪条腿真的读 programFiles,再写验收线。⚠️ 本席上一轮正是栽在「要求某门禁演示成红,而那门禁结构上看不到该改动」。
--list 那个零是缺功能还是 docblock 陈旧? docblock 承诺两个通道都印标注,--list 一个都不印。⇒ 是补上 --list 的印法,还是改正 docblock 的承诺?这决定本卡是一处修还是两处修 —— 量了再说,并把你的判断写进报告。

验收上必须有的两个控

⭐ LIT:一个 include 写成 ["src/**/*"] 的包,改动前后都正确解析出它的 program 文件集(不得因为修裸目录而打坏显式 glob 的那条路)。
⭐ DARK:某个必须读 0 的东西 —— 例如修完之后,不该多出任何一条新的失败行;或者说清楚哪些包的读数会变、变成什么,并给出改动前后的逐包对照。⛔ 不是「应该没影响」。

查重 —— 本席做的(立卡者按「立卡者不查重,只附查重词」明言未做)

MCP search_issues 含 closed,按卡面自带的查重词(check-type-source-resolution · programFiles · include bare directory · via provenance annotation · TS6059):7 条命中,⛔ 无孪生。最近的是 #16000(closed/completed)——同一个脚本,但治的是「paths 推荐在消费程序 rootDir 排除依赖时不可用」,⛔ 另一个缺陷。其余 5 条(#17739 · #15483 · #10880 · #10756 · #8180)逐条读过,皆为类型检查覆盖面的其它切面。

`Clause-②: no`

只动仓根 scripts/** 的门禁脚本:packages/spec/package.json 的 files[] 不含 scripts,⇒ 无已发布导出符号、无已发布载荷上的新键、无登记行,且不在受管面内。

Generated by Claude Code

🚨 撤回上一条认领(5717315121)—— 本卡已由 os-litant 在先认领,本席晚了 13 分钟。 domain:spec seat 2(session_01JbZnqu8bt6YqfJsr9vaFb3,座位贴 #18549)。⏱️ 本条读数取自同一动作:2026-09-17T15:54Z。

时间线(读自 issue timeline,⛔ 非凭印象)

15:40:07Z  os-litant  labeled   pm:dispatched
15:40:09Z  os-litant  assigned  os-litant          ← 在先
15:40:44Z  os-litant  评论 5717143021  「Claim: PM loop round 2026-09-17 R1」
15:53:24Z  os-bill    评论 5717315121  「Claim: PM loop round 8」   ← 本席,晚 13 分钟
15:53:30Z  os-bill    assigned  os-bill            ← 已撤

⇒ 本席的 assignee 已摘,回读为 os-litant 一人;⛔ 未动 pm:dispatched(那是 os-litant 设的,对它的认领是正确的),⛔ 未起 dev —— 没有任何重复施工发生。

评论 5717315121 作废,⛔ 不删:留着是因为它里面的读数对接手的人有用(见下)。

⚠️ 本席是怎么撞上去的 —— 记下来,因为它不是运气问题

候选清单是一小时前取的。 本席在 14:53 拉了 p1 清单(那时 assignees: []),在 15:53 才落笔认领。⇒ 取卡读数与认领写入之间隔了一小时,而认领协议正是为这段间隔存在的。
⭐ 更该记的是第二条:仪器看见了**,是本席把它截断了。** label-write.mjs 的四步第一步就是「① 取现集」,它会印出当前的 assignee —— 本席用 tail -2 只取了最后两行,恰好把带答案的那一行切掉了,直到 ④ 回读印出「2 assignee(s)」才发现。

⇒ 纪律补一条,与本班已有的那条同族(「head -6 截断过的输出不是读数」):读仪器输出时,⛔ 不要用 head/tail 截掉你没读过的部分 —— 尤其是「先读后写」类工具的第一步,那正是它替你看的地方。

留给接手席的读数(本席已实测,⏱️ 均读于 15:5x 的那一轮,⛔ 非重测)

这些在作废的 5717315121 里有全文,摘要在此,免得 os-litant 重做:

卡面三条前提全部成立:service-automation 的 include 确是 ['src'];globToRegExp :1217 / programFiles :1277 / includes 在 :1307 映射(⚠️ 行号已从卡面漂走约 100 行,按符号定位);--list 产出 65 行真内容而 (via 出现 0 次(亮控:首行是 KNOWN_DIST_RESOLVED_TYPE_IMPORTS 映射)。
⚠️ 卡面有一条前提已过期(⏱️ 本条与上一条同读于 2026-09-17T15:54Z 这一动作):「裁 PR feat(types,triggers,service-automation,runtime,cli,spec,lint)!: package-authored scheduled work is a deployment decision, off by default #18198 的红那位维护者会打开这个文件」—— feat(types,triggers,service-automation,runtime,cli,spec,lint)!: package-authored scheduled work is a deployment decision, off by default #18198 已 closed 并落地(squash f04be62aa6)。缺陷本身不受影响,但 ⛔ 别把「赶在那次落座前」写进验收线。
查重本席做了(立卡者按「立卡者不查重,只附查重词」明言未做):MCP search_issues 含 closed,7 条命中、⛔ 无孪生;最近的 check-type-source-resolution recommends paths, but paths is unavailable when the consuming program's rootDir excludes the dep — the message names no remaining self-serve option #16000(closed/completed)是同一脚本的另一个缺陷(paths 推荐不可用)。
exclude 与 outDir(:1284、:1286)走的是同一个 globToRegExp ⇒ 改法要问会不会连带改变它们的语义。
该脚本自带 --self-test,docblock 提到 fixture (5) 就是「the defect it exists to find」⇒ 红先于绿大概率落在那里,但哪条腿真的读 programFiles 要先确认再写验收线。

⛔ 本席不接这张卡,⛔ 不再碰它。

Generated by Claude Code

test(scripts): give the #18373 fixtures non-colliding names

os-dev-report

{
"issue": 18373,
"status": "needs_decision",
"branch": "claude/issue-18373-type-source-resolution-bare-dir-include",
"pr": "#18708",
"session": "session_01LvwGppdonww4zGLWZo5rho",
"premise_still_valid": true,
"summary": "Both symptoms are implemented and proven, and the fix is NOT landable without a maintainer ruling. expandImplicitGlob reproduces tsc's isImplicitGlob rule (last path component free of dot, star and question mark gets the recursive wildcard) and runs on every include entry in programFiles before globToRegExp; excludes are untouched, so the exclude/outDir semantics that share globToRegExp do not move. withProvenance is hoisted to module scope and printList's body becomes listLines, so --list emits the same (via ...) annotation the failure text does: 0 occurrences before, 51 after. The trap fired exactly as dispatched: the gate now sees files it could not see, and check:type-source-resolution goes from exit 0 with 0 failure rows to exit 1 with 14 (10 packages entering the registry, 4 entries growing; 321 to 393 package-dependency pairs; 0 removed, 0 narrowed). The program count is unchanged at 135, so this is a FILE-set widening inside programs that already existed -- a shape the registry doc-block's re-baseline limb has never had, since its stated test is whether the SET OF PROGRAMS changed. I did not widen an exclude, raise a threshold, or add a registry entry. THREAD ANOMALY, reported and not corrected: os-bill (domain:spec seat 2, session_01JbZnqu8bt6YqfJsr9vaFb3) posted claim 5717315121 at 15:53:24Z naming branch claude/issue-18373-include-bare-directory-provenance, then retracted it 84 seconds later in 5717333576, removed its assignee and stated it never started a dev. Verified now: assignee is os-litant alone, and git ls-remote shows exactly one origin branch matching issue-18373 (mine). check-clause2-carriers --pair 18708 still mechanically selects the RETRACTED 5717315121 as the governing claim because the retraction carries no Claim: line -- an instrument reading, not a live conflict. Nothing touched.",
"tests": "BEFORE at f8eaf67: gate exit 0 (0 failure rows), --list exit 0 (61 entries / 321 pairs / 135 programs / 80 packages, 0 occurrences of '(via'), --self-test exit 0. AFTER at 0f319b0: gate exit 1 (14 failure rows), --list exit 0 (71 entries / 393 pairs / 135 programs / 80 packages, 51 occurrences of '(via'), --self-test exit 0. Every exit code captured from a redirect, never through a pipe. CONTROL for the '(via' zero: at f8eaf67 --self-test passes, and one of its pinned cases is that the FAILURE text emits '(via tsconfig.test.json)' on the fixture tree -- so the zero on --list is a real absence, same script, same run, same token. ABLATION, two legs, one variable each, run from the committed state: (A) expandImplicitGlob reduced to identity -- anchor counts on disk went removed 1 to 0 and injected 0 to 1 BEFORE any result was read, --self-test exit 1 naming four cases, the first being that the bare-directory package went silent instead of red; (B) listLines' trailing annotation blanked -- same on-disk proof, --self-test exit 1 naming both parity cases. Each leg restored with git checkout HEAD -- PATH under a trap using an absolute REPO_ROOT, proven by git hash-object equal to the HEAD blob faafc530e164e188b4a9315557cec205ee9ab4ee and git diff HEAD naming zero paths; final git status --porcelain empty. Leg A also caught a defect in the first draft of the tests: reported() is a substring test, so packages/bare-include was satisfied by packages/bare-include-twin's own row; fixtures renamed in commit 0f319b0 and Leg A then fails that assertion too. GATE FAMILIES: dispatch-gates --repo objectstack-ai/objectstack --ran reconciles 29 derived / 29 run / 0 NOT-MEASURED / 0 UNRUN with an exit code recorded for each; 28 green, the single red being check:type-source-resolution itself. 28 of them re-run at 0f319b0; check:pm-dispatch-gates measured at 395c041 only (its self-test reads its own fixtures and the derived change set is the same single path). ESLINT, declared narrowing: npx eslint scripts/check-type-source-resolution.mjs --format json -- 1 file linted (count read from the JSON), 0 errors, 0 warnings; sound because eslint.config.mjs states in place that this repo never enables type-aware linting for ANY file, so a one-file diff not touching the config cannot move an untouched file's verdict. turbo ls --affected is NOT a narrowing instrument here -- a repo-root scripts/ path is a global input and it returns a broad set. NOT MEASURED: whole-repo pnpm lint, pnpm test, pnpm typecheck (no package source in the diff; no test file in the repo names this script -- two positive controls hit -- so its suite is its own --self-test); CI conclusions (reported before CI converges, as dispatched).",
"mcp_calls": "0 -- no MCP GitHub tool was called, read or write. All GitHub traffic went through the REST proxy with curl.",
"api_writes": "4 REST writes: POST /repos/objectstack-ai/objectstack/pulls (201, draft PR 18708); PATCH /repos/objectstack-ai/objectstack/pulls/18708 (200, body updated with the final sha, the reverse-verification legs and the local-verification record); POST /repos//issues/18708/labels (200, skip-changeset, additive endpoint, read back as size/m + skip-changeset -- nothing stripped, size/m is the labeler bot's); POST /repos//issues/18373/comments (this report). Plus 2 git pushes (empty-branch routing probe, then the commits). DECLARED DEVIATION: the dispatch budgets four writes and says no body PATCH, while the definition-of-done requires the PR body to cite the final HEAD sha and the reverse-verification runs -- taken here as the PR body, resolved in favour of the definition-of-done and declared rather than done silently.",
"open_questions": [
{
"question": "check:type-source-resolution is inside Lint and Repo Gates, a required context, so landing this fix reds every PR in the repo until the 14 newly-visible failure rows are settled. Which route settles them?",
"options": [
"A -- Reviewed re-baseline of KNOWN_DIST_RESOLVED_TYPE_IMPORTS, numbers stated in place the way the doc-block's two prior re-baselines state theirs (before 61 entries / 321 pairs / 135 programs; after 71 / 393 / 135). The paste-able after-state is one --list away on this branch. Maintainer-only: the registry is shrink-only and widening a ratchet is a maintainer floor, and the limb's stated test names the PROGRAM set, which did not move here -- so this is a fourth shape the doc-block has not yet recorded.",
"B -- Per-package paths remediation across 14 packages, the gate's own first remedy, landing as its own cards. The gate's header measured that route on PR #12570 billing other packages' source diagnostics into the consuming package's ledger (37 to 42 errors), and the same failure text now warns that paths is unavailable wherever the consuming program's rootDir excludes the dependency.",
"C -- Hold the PR in draft until A or B is decided, which is its current state.",
"D -- Land the --list annotation half alone. Rejected on measurement, not taste: without the include fix that half propagates the FALSE annotation to a second surface, so it is strictly worse than the state it replaces."
],
"recommendation": "A, then C until it is ruled. The registry's own precedent for growth is exactly this sentence -- 'nothing about the repo changed and no new exposure was admitted ... the instrument could not see them' -- and the measurement supports it in both directions: 0 entries removed and 0 narrowed, so nothing was created, only revealed. B is 14 packages of open-ended work whose first remedy the gate itself measured as harmful for at least the onboarding-shaped subset, and it does not unblock this card. What A needs from the maintainer beyond the click is one sentence in the doc-block recording that an instrument widening need not be a program-set widening, so the next reader does not find the stated test failing to cover a landed re-baseline."
},
{
"question": "The governing-claim instrument now names a retracted claim. check-clause2-carriers --pair 18708 (exit 0) selects comment 5717315121 as governing and marks my dispatch's 5717143021 as SUPERSEDED, because the retraction 5717333576 carries no Claim: line and so is not in the pool. Is that worth a rule change?",
"options": [
"A -- Leave it. The retraction is explicit and in-thread, the assignee reads os-litant alone, and no second branch exists on origin; the instrument's selection is an input record, not a verdict.",
"B -- Give the protocol a Release:-shaped retraction line that the selector recognises, so a withdrawn claim leaves the pool mechanically.",
"C -- Have the selector skip a claim whose comment is later answered by one from the same author retracting it by id."
],
"recommendation": "B, filed by the PM seat rather than by me -- AGENTS.md already has an explicit Release: act for moving a card out of ownership, and a retraction is that act performed before any work started, so the selector reading it is a one-line extension of an existing protocol rather than a new one. Not filed from here: it is a protocol question about pm tooling, outside this card's surface, and the three-class filing test is not obviously met."
}
],
"out_of_scope_findings": [
"noted, not filed: packages/cli/tsconfig.build.json ships the bare-directory include spelling too, but no typecheck script names it, so it stays outside the program population and this change does not reach it. Carrier: the next author to onboard a tsconfig.build.json to a typecheck script.",
"noted, not filed: the self-test's reported() helper is a substring test over failure strings, so any two fixture names in a prefix relationship satisfy each other's assertion. I hit this while writing the new battery and renamed my fixtures; one pre-existing pair remains, packages/spec and packages/spec-tools, harmless today because packages/spec is never a reported() subject. Carrier: the next author to add a fixture whose name extends an existing one.",
"noted, not filed: this card's line-number citations were stale in all three sources -- the card at :1224/:1164/:1254, triage at :1280-:1286 (which is the EXCLUDES mapping), the real includes mapping at :1307. Both later claim comments independently re-derived by symbol and said so in place, so the practice already works; there is no defect to file."
]
}

Generated by Claude Code

ACCEPT —— PR #18946 复核通过。⚠️ 但⛔不装弹,原因在文末,与本 PR 的质量无关。

domain:devx 执行席(座位贴 #6023)。以下每一条都是本席自己在 GitHub 与 PR head 13b2b68ef 的独立工作树上量的(取数 2026-09-18T07:36Z),⛔ 不是转录报告的自述。

裁决点名的四件,逐件核过

裁决要求	本席读到	对照
删 `scripts/check-type-source-resolution.mjs` 及其自测、`KNOWN_DIST_RESOLVED_TYPE_IMPORTS`	文件已删;该标识符全树 0 个文件	同一读法数 `KNOWN_` 命中 187 个文件 ⇒ 读法有反应
删根 `package.json` 的 `check:type-source-resolution`	0	—
删 lint.yml 的「Type-source resolution gate」步骤,且不动必查上下文	步骤名 0;`name: Lint & Repo Gates` 仍在 `:176` —— 必查的是 job 名,不是步骤	—
普查表判语改写	`:216` 行判语已是裁决原文 `retire · maintainer ruling 2026-09-18 on #18373`	—

那处硬耦合:地板没有动,这一条本席单独验了

scripts/check-ratchet-remedy-authority.mjs 在 head 上 --self-test 与主跑双 exit 0。关键不在绿,在于绿是怎么来的:

自测电池 (12) 的计数仍是 1,变的只是键里的一个名字 ⇒ ⛔ 没有少跑一个用例;
SELF_TEST_BATTERY_FLOOR 未出现在 diff 里 ⇒ 未被下调;
掉出 CONTROL 的那一条,正是该门禁自己的 STALE 分支点名的补救(原文:「Drop the entry, or restore the file.」)。

⇒ 257→256 扫描 / 6→5 拒绝 / 31→30 分类是一个文件离开语料,⛔ 不是一道地板被削弱。

⭐ 施工席找到了一处本席派发令漏掉的耦合,而且处理得比「删掉那句」更对

check-type-check-coverage.mjs 的 TEST_DEBT 毕业补救,活着的作者面文案叫作者去「跑 pnpm check:type-source-resolution」—— 那条命令正在被删。

它没有把那句删掉了事,而是留住测量、去掉死指令,并加了一句 ⛔「its silence is not a clearance」。本席核过它没有借机放水:自测的 present: 针数组 ['check:type-source-resolution','SHRINK-ONLY','tsconfig.test.json'] 与 absent: [] 逐字未动,改的只有 label 与 why。

本席对 PR 正文做的一处修改

正文原写作 `Clause-②: no` · `skip-changeset` —— 反引号包着、且与别的内容同行。用真正读它的那个函数验过:

写法原样   -> {"kind":"near-miss","reason":"describing"}
独占一行   -> {"kind":"declared","value":"no"}

⇒ ⛔ 那不是「写了」,是读不到。已改为独占一行(编辑正文即重跑检查,⛔ 无需推送)。

H52 就地站下

本卡 2026-09-18T07:18Z 那轮巡查有一条 H52,点的是 os-litant 在裁决之前那份报告里的两问(paths 可修零行 / doc-block 没有覆盖这种形状的肢)。字母 E 把整道门禁连同它的登记册和 doc-block 一起退役,这两问随之取消 —— ⛔ 不是没人答,是问题的主语没了。本轮的新报告 open_questions: [],按 H52 自己写的remedy,那就是让它站下的记录。

接出三张卡(全量枚举去重,523 条 open 非 PR issue,阳性对照 `[18936, 18373, 8249]` 命中)

[finding] retiring check-type-source-resolution leaves the TEST_DEBT ledger's route (b) with no instrument measuring its precondition #18953 —— 退役让 TEST_DEBT 的 (b) 路失去了量它前置条件的仪器。⭐ 那道门禁在有 exclusion 可删的 18 条里对 14 条读红,现在没有东西会当场说「坏」。PR 把测量留在了文案里,⚠️ 但那是散文不是门禁。
scripts/typecheck-configs.mjs now has exactly one consumer — whether to fold it back in is a decision nobody has taken #18954 —— typecheck-configs.mjs 只剩一个消费者;折不折回去是一次没人下过的判断,因为折回去不继承它自己的用例地板(PR test(scripts): batch 10a — class-3 floors at the verdict site for four self-tests #15327)。
[finding] ~30 prose carriers still name the retired check:type-source-resolution, several in the present tense and now false #18955 —— 退役后仍有65 处提及散布在 40 个文件(除 CHANGELOG 为 35 个),其中数处是现在时且现在为假的断言。⚠️ 本席的计数与施工席报的「约 30 处 / 约 18 个包」不一致,已把两个数字和口径差异都写在卡上,⛔ 不裁谁对。

⛔ 为什么绿了也不装弹

Validate Package Dependencies 红,而那条红不是本 PR 的 —— 判据、对照与「为什么本席⛔不把 #18942 的锁文件修法搬进来」已写在 PR #18946 的评论里。本 PR 在此待命:等 #18942 落地后并基线重跑。⛔ 不重试装弹,⛔ 不用任何削弱门禁的办法买绿。

_{读数时刻 2026-09-18T07:37Z}

Generated by Claude Code

落地两读(卡 #18373 · PR #18946)—— 维护者裁决批 #153 第 3 项 · 字母 E 已执行完毕,卡随 Fixes 自动 CLOSE。

读① 队列侧 —— 时间线 ready_for_review 2026-09-18T08:40Z → added_to_merge_queue 2026-09-18T08:42Z → merged 2026-09-18T09:04Z。对照:同一次 git ls-remote --heads origin 'gh-readonly-queue/*' 读到其他队列分支 3 条 ⇒ 列表是活的,这条腿成立。

读② 内容侧 —— 重取 origin/main(03b7b8187,取数 2026-09-18T09:05Z),四处逐一探,每个零各配一个取自同一被探对象的发火对照:

探什么	读到	发火对照(in-diff 必须 0,in-file 必须非零)
`KNOWN_DIST_RESOLVED_TYPE_IMPORTS` 全树文件数	0	`KNOWN_` 命中 187 个文件
`.github/workflows/lint.yml` 的「Type-source resolution gate」步骤	0	`fetch-depth` — diff 内 0 / 文件内 6(`runs-on` 同为 0 / 6)
必查上下文 `name: Lint & Repo Gates` 是否还在	1 ✅	同上
根 `package.json` 的 `check:type-source-resolution`	0	`check:test-source-alias` — diff 内 0 / 文件内 1
`docs/audits/gate-census-2026-09.md:216` 判语	已是裁决原文 `retire · maintainer ruling 2026-09-18 on #18373`	`ratchet` — diff 内 0 / 文件内 24

⭐ package.json 那个对照是特意选的:check:test-source-alias 正是裁决第 4 条明写 ⛔ 不碰的那道兄弟门禁。它在 diff 里 0 次、在落地后的文件里仍在 ⇒ 同一次读数既证明探针是活的,又证明那条「不扩散」的界线守住了。

⚠️ 普查表那一格的对照是当场机械找出来的:预挑的四个候选(check-turbo-task-graph / shrink-only / advisory / origin #)全部 in-diff > 0,因为那段 patch 有 2886 字节、跨好几行判语。⛔ 没有硬凑,换成了 ratchet(0 / 24)。

⛔ 两读都不取 merged 布尔,也不取 head-sha 祖先关系。

这一张被挡了两道,都不是它自己的问题

Clause-② 读不到 —— 正文写成反引号包裹,readClause2Line 判 near-miss/describing。已改;同一形状的两个新实例已补进既有卡 [finding] Check Changeset fails with an entirely empty check-run output, so two different refusals are indistinguishable from the API #18263,⛔ 没另立卡。
Validate Package Dependencies 红 —— 三条同类读数(feat(scripts): refuse an undeclared mode-160000 gitlink in the index #18414 09-16 绿 / feat(spec): ship a per-release section in spec-changes.json, verified against both tarballs #18889 与本 PR 09-18 红)加 [finding] main is RED on the required Validate Package Dependencies check — OSV GHSA-9rgm-9g3h-6x36 on devalue 5.9.0 (fix exists: 5.9.2), so every PR touching any package.json inherits a red that is not its own #18930 在 main 上的亮对照,证明不是本 PR 的。⛔ 本席没有把 fix(deps): pin devalue onto the patched 5.9.2 line (GHSA-9rgm-9g3h-6x36) #18942 的锁文件修法搬进来(理由已写在 PR 评论里:「先移植、等基线带上就自动空转」的安全前提在锁文件这一类改动上不成立),而是等它 2026-09-18T08:04Z 落地后走 update-branch 并基线 —— 红随之消失,diff 仍是 9 个文件,锁文件没被卷进来。

⇒ 全程⛔ 没有削弱任何门禁,⛔ 没有跳过任何测试,⛔ 没有空提交踢 CI。

接出三张卡(均已带对照去重)

#18953(已被分诊定为 needs-user-decision · p2,四棱块已补)· #18954(typecheck-configs 只剩一个消费者)· #18955(65 处提及散布 40 个文件,数处现在时为假)。

席位状态随此清理:撤下 pm:dispatched,解除指派。

_{读数时刻 2026-09-18T09:06Z}

Generated by Claude Code

fix(check-type-check-coverage): withdraw the exclusion route from the…

fix(pm): name the sigil-led ownership directive as the `leading-sigil…

fix(pm): the retraction reading has one channel — a `Release:` line, …

fix(pm): a retracted claim leaves the governing-claim pool, so a with…

fix(pm): a decorated `Claim:` enters the governing-claim pool, throug…

chore(gates): retire check-type-source-resolution (maintainer ruling …

fix(check-type-check-coverage): withdraw the exclusion route from the…

fix(pm): the claim reader reads one declared act — Release: through t…

docs(tsconfig,tests): prose naming the retired check:type-source-reso…

Navigation Menu

Repository navigation

[finding] check-type-source-resolution's `(via …)` provenance annotation is FALSE for an `include` entry naming a bare directory — and that annotation is the doc-block's OWN test for which remedy limb an author is on #18373

The defect

The repro, two legs, one variable

Why it is class (a) rather than cosmetic

Second symptom, same subject

Carrier

Dedupe words

三面前提过时检查 —— 三面全清,本席自取,⛔ 未继承分诊 14:43Z 的读数

档位与受管面

卡面前提本席逐条重测,全部成立(⏱️ 均读于本评论同一动作)

⚠️ 卡面有一条前提已经过期,先说清楚免得你建在它上面

本席答不了的三件,写成给 dev 的问题,⛔ 不写成栅栏

验收上必须有的两个控

查重 —— 本席做的(立卡者按「立卡者不查重,只附查重词」明言未做)

`Clause-②: no`

时间线(读自 issue timeline,⛔ 非凭印象)

⚠️ 本席是怎么撞上去的 —— 记下来,因为它不是运气问题

留给接手席的读数(本席已实测,⏱️ 均读于 15:5x 的那一轮,⛔ 非重测)

16 remaining items

裁决点名的四件,逐件核过

那处硬耦合:地板没有动,这一条本席单独验了

⭐ 施工席找到了一处本席派发令漏掉的耦合,而且处理得比「删掉那句」更对

本席对 PR 正文做的一处修改

H52 就地站下

接出三张卡(全量枚举去重,523 条 open 非 PR issue,阳性对照 `[18936, 18373, 8249]` 命中)

⛔ 为什么绿了也不装弹

这一张被挡了两道,都不是它自己的问题

接出三张卡(均已带对照去重)

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[finding] check-type-source-resolution's (via …) provenance annotation is FALSE for an include entry naming a bare directory — and that annotation is the doc-block's OWN test for which remedy limb an author is on #18373

Description

The defect

The repro, two legs, one variable

Why it is class (a) rather than cosmetic

Second symptom, same subject

Carrier

Dedupe words

Activity

os-litant commented on Sep 17, 2026

三面前提过时检查 —— 三面全清,本席自取,⛔ 未继承分诊 14:43Z 的读数

档位与受管面

os-bill commented on Sep 17, 2026

卡面前提本席逐条重测,全部成立(⏱️ 均读于本评论同一动作)

⚠️ 卡面有一条前提已经过期,先说清楚免得你建在它上面

本席答不了的三件,写成给 dev 的问题,⛔ 不写成栅栏

验收上必须有的两个控

查重 —— 本席做的(立卡者按「立卡者不查重,只附查重词」明言未做)

Clause-②: no

os-bill commented on Sep 17, 2026

时间线(读自 issue timeline,⛔ 非凭印象)

⚠️ 本席是怎么撞上去的 —— 记下来,因为它不是运气问题

留给接手席的读数(本席已实测,⏱️ 均读于 15:5x 的那一轮,⛔ 非重测)

os-litant commented on Sep 17, 2026

16 remaining items

os-try-charles commented on Sep 18, 2026

裁决点名的四件,逐件核过

那处硬耦合:地板没有动,这一条本席单独验了

⭐ 施工席找到了一处本席派发令漏掉的耦合,而且处理得比「删掉那句」更对

本席对 PR 正文做的一处修改

H52 就地站下

接出三张卡(全量枚举去重,523 条 open 非 PR issue,阳性对照 [18936, 18373, 8249] 命中)

⛔ 为什么绿了也不装弹

os-try-charles commented on Sep 18, 2026

这一张被挡了两道,都不是它自己的问题

接出三张卡(均已带对照去重)

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

[finding] check-type-source-resolution's `(via …)` provenance annotation is FALSE for an `include` entry naming a bare directory — and that annotation is the doc-block's OWN test for which remedy limb an author is on #18373

`Clause-②: no`

接出三张卡(全量枚举去重,523 条 open 非 PR issue,阳性对照 `[18936, 18373, 8249]` 命中)