pm(labels): stop creating needs:contract-review and drop its last code legs

ensure-pm-labels.sh no longer creates the label (the GitHub object is the
seat's to remove); check-changeset-no-major reads the Clause-② line alone —
the label carrier axis, the check-half-states import and the #17229 battery
(floor 20 → 19) go; check-clause2-carriers.mjs keeps a LOCAL retired constant
because its retired --pair sweep still compares it and the file cannot be
deleted yet (the queue guard lazily imports its record recognisers); the
label-desc-cap specimen, the workflow comments, the CONTRACT_REVIEW_TIER
docblock and post-stamped's near-miss row stop naming the label or --pair.

Claude-Session: https://claude.ai/code/session_01Wnstp2kTth7sGXfr8fXypc
Co-authored-by: Claude <noreply@anthropic.com>

Author: claude

.github/workflows/lint.yml

@@ -916,9 +916,9 @@ jobs:
       # over-long -d means the label is never created on a repo that lacks it
       # and a rerun can never repair it. Headroom is thin (one description sits
       # at exactly 100), and the only previous enforcement was the author
-      # counting by hand. Measured in characters, never bytes: the live
-      # needs:contract-review description is 97 characters / 101 bytes, so a
-      # byte-length guard would red a label GitHub accepts. Unconditional, like
+      # counting by hand. Measured in characters, never bytes: the specimen
+      # that fixed the measure was 97 characters / 101 bytes, so a byte-length
+      # guard would red a label GitHub accepts. Unconditional, like
       # the other self-tested gates above.
       - name: PM label description cap
         run: pnpm check:pm-label-desc-cap
@@ -1254,16 +1254,15 @@ jobs:
       # else's pair is a fact about the board, not about the PR that happens to
       # run CI next, and failing that PR would punish the wrong actor.
       #
-      # ⛔ Do not "promote" the live sweep into this step. The script's own
-      # `--pair <pr>` mode is the per-PR predicate, and it is a PRE-ARM tool a
-      # seat runs against its own pair — deliberately not a merge gate, because
-      # a new blocking context over board state is a required-set change and
-      # the required set is the maintainer's.
+      # ⛔ Do not "promote" the live sweep into this step. The script's `--pair`
+      # landing pre-check is retired (ruling record 5770886272 on #19061) and
+      # nothing runs it; the file stays only because the Governed Surface Queue
+      # Guard lazily imports its contract-review record recognisers and
+      # `--template`, and this self-test is what keeps those loadable.
       #
       # The self-test is offline and imports `check-half-states.mjs` for the
-      # label constant, the `prDeliversCard` delivery relation and the shared
-      # PREREQUISITE-NOT-MET exit code — one source for each, so this checker
-      # and H31 cannot come to disagree about which PR delivers which card.
+      # `prDeliversCard` delivery relation and the shared PREREQUISITE-NOT-MET
+      # exit code — one source for each.
       - name: Clause-② carrier checker self-test
         run: pnpm check:pm-clause2-carriers
 

.github/workflows/pr-automation.yml

@@ -784,8 +784,8 @@ jobs:
                WHAT THE AXIS DOES NOT JUDGE, in its own words: "WHICH of the
                packages above received that widening is NOT something this gate
                can read, and it does not claim to. Clause ② is declared ONCE, FOR
-               THE PR" -- neither the carrier label nor the declaration line in
-               the PR body names a package. The axis cross-checks a declaration
+               THE PR" -- the declaration line in the PR body names no package.
+               The axis cross-checks a declaration
                against the levels; it never reads the diff to find the act.
                Judging the act is still yours, this paragraph is still where the
                rule lives, and everything in it the axis does not reach is still

scripts/check-changeset-no-major.mjs

@@ -710,7 +710,6 @@ import { isEntrypoint } from '../invoked-as.mjs';
 import {
   CLAIM_COMMENT_MARKER,
   CONTRACT_REVIEW_HEADING_MARKER,
-  CONTRACT_REVIEW_LABEL,
   DEFAULT_SWEEP_REPO,
   EXIT_PREREQUISITE_NOT_MET,
   H51_SHA_MIN_HEX,
@@ -726,7 +725,6 @@ import {
   claimGovernance,
   claimedBranches,
   governingClaim,
-  isGateSemanticLabel,
   labelNames,
   latestMarkedComment,
   markerMatches,
@@ -756,6 +754,23 @@ import { CONTRACT_REVIEW_TIER } from './dispatch-gates.mjs';
 // this file is the SWEEP around them, never a second reading of them.
 import { CLAUSE2_ARMS, CLAUSE2_VALUES, quoteLine, readClause2Line } from './clause2-line.mjs';
 
+/**
+ * ⛔ RETIRED — the clause-② gate label and its dual-carrier discipline left the
+ * tree with ruling record 5770886272 on #19061 (letter B): `ensure-pm-labels.sh`
+ * no longer creates it, the queue guard's label leg is gone, and the half-state
+ * rows that patrolled it (H31/H35/H51/H53/H61) are deleted. The constant lives
+ * HERE, in the one file that still reads it, because this file's `--pair` sweep
+ * — itself retired by the same ruling, and run by nothing — compares the label
+ * on its C1/C3 rows and the self-test pins that comparison. The file survives
+ * only until the record recognisers and `--template` the queue guard imports
+ * (`locateReviewOfRecord`, `deliveredCardNumber`, `servedTierStands`,
+ * `isModelIdentifierToken`, `REVIEW_OF_RECORD_THREADS`,
+ * `REVIEW_OF_RECORD_LOCATION`) are extracted into a module of their own, and
+ * this constant goes with it. ⛔ Nothing new may import it.
+ */
+const CONTRACT_REVIEW_LABEL = 'needs:contract-review';
+const isGateSemanticLabel = (name) => String(name ?? '') === CONTRACT_REVIEW_LABEL;
+
 /**
  * The lanes that OWE a review of record on EVERY round they deliver -- the
  * maintainer's lane rule, restated on #18536 and carried into

scripts/pm/check-clause2-carriers.mjs

@@ -28,9 +28,9 @@
  * ## CHARACTERS, not bytes — the measure is the whole point
  *
  * The cap is on characters, and the descriptions in this file are not ASCII:
- * they carry em dashes, circled digits and CJK. The live specimen is
- * `needs:contract-review`, whose description is 97 characters and 101 BYTES —
- * accepted by GitHub, and red under a naive byte-length guard. A byte guard
+ * they carry em dashes, circled digits and CJK. The specimen that fixed the
+ * measure was a label description of 97 characters and 101 BYTES — accepted
+ * by GitHub, and red under a naive byte-length guard. A byte guard
  * would therefore fail a label that works, and the fix a reader would reach for
  * (rewriting a correct description to please the gate) is worse than no gate.
  *
@@ -379,9 +379,10 @@ function selfTest() {
     }
   };
 
-  // The measure: characters, not bytes, not UTF-16 units.
+  // The measure: characters, not bytes, not UTF-16 units. The specimen is the
+  // retired gate label's description, kept as a fixture for its byte/char gap.
   const live = 'Clause-② enqueue gate: dispatched below contract-review tier — blocked until the review clears it';
-  t('the live 97-char/101-byte description measures 97', charLength(live), 97);
+  t('the 97-char/101-byte specimen measures 97', charLength(live), 97);
   t('…and is over the cap only if you measure BYTES', Buffer.byteLength(live, 'utf8') > MAX_DESCRIPTION_CHARS, true);
   t('…so it passes this gate', violations([{ label: 'x', line: 1, expanded: live, unresolved: [] }]).length, 0);
   t('an astral character counts once, not twice', charLength('🚀'), 1);

scripts/pm/check-label-desc-cap.mjs

@@ -161,8 +161,8 @@
  * contract this is.
  *
  * **It writes nothing and hangs no label.** Same call `check-clause2-carriers`
- * and `check-half-states` make: a checker that hung `needs:contract-review`
- * would be issuing the review verdict, which is 自查放行. ⛔ No new label and no
+ * and `check-half-states` make: a checker that hung a review-gate label would
+ * be issuing the review verdict, which is 自查放行. ⛔ No new label and no
  * new claim-line syntax exist because of this file — #16448 forbids both, and
  * the reader it uses is the sibling's existing `Clause-②:` reader.
  *

scripts/pm/check-widening-tells.mjs

@@ -30,13 +30,13 @@
  * `pnpm install`, and both gates stage their own first-party closure into a
  * throwaway repo for their I1/I2 fixtures.
  *
- * ⚠️ This file is STEP ① of that ruling and nothing else of it has landed yet:
- * the `needs:contract-review` label, `check-clause2-carriers.mjs --pair`, the
- * double-carrier discipline and the independence pair are retired BY THE RULING
- * and still RUN in this tree — the label exists, `--pair` still judges its C4 /
- * C6 / C7 rows, and `check-governed-merges.mjs` still prescribes `--pair N` at
- * 0 — until step ② (the rule text) and step ③ (the label) land. ⛔ Read the
- * paragraph above as what was DECIDED, never as what this tree already does.
+ * This file was STEP ① of that ruling; step ② retired the rule text and step ③
+ * the label — its constant, the queue guard's label leg, the half-state rows
+ * that patrolled it and the changeset gate's label carrier are gone from this
+ * tree. What still stands beside this reader is `check-clause2-carriers.mjs`,
+ * kept only because the queue guard's Tier S leg lazily imports its record
+ * recognisers and `--template`; its `--pair` sweep is retired by the ruling and
+ * nothing runs it, and the file goes once those recognisers are extracted.
  *
  * ## ⛔ One reader, and this is it
  *

scripts/pm/clause2-line.mjs

@@ -12228,9 +12228,9 @@ export function residueLines(
  * reads — clause ②'s CONTRACT-REVIEW tier: the tier the clause-② REVIEW runs
  * at, both halves of it — the spec and skills lanes' review of every round
  * they deliver (a card that changes contract accept/reject behaviour or
- * widens the public surface is spec-lane work, whichever seat found it), and
- * the `needs:contract-review` re-review sub-round (its opening self-check
- * reads this). The BUILD of such a card is at the default judgment tier, so
+ * widens the public surface is spec-lane work, whichever seat found it) and
+ * the `Served-tier:` line every `## Contract review` record carries. The
+ * BUILD of such a card is at the default judgment tier, so
  * this constant is a review tier and never a dispatch mandate. Declared HERE
  * and only here, as a constant, so a model upgrade is a one-line change in one
  * file — the clause-① mandate rows below read it, the self-test compares
@@ -12251,7 +12251,7 @@ export function residueLines(
  * and never self-reviewed. A tier that has been RETIRED never comes back, and
  * the two cases differ on WHO MAY ACT: a seat reading 「⛔ 不许降档」 onto a
  * vanished tier holds its whole lane forever, and a seat picking the
- * replacement itself is the silent downgrade the fuse exists to stop. So a
+ * replacement itself is the silent downgrade this rule exists to stop. So a
  * retirement is a maintainer ruling and ⛔ never a seat's reading — and when
  * the ruling lands, this VALUE is the one line that moves. The ceiling of the
  * ladder `tierLines` prints is DERIVED from it ({@link TIER_CEILING}) so the
@@ -12277,7 +12277,7 @@ export function residueLines(
  * that session; this constant is a property of the lane's governance, and the
  * two ⛔ never trade places.
  *
- * Rulebook: `.claude/skills/pm-dispatch/SKILL.md` 「入队与落地」 — the clause-② gate and the `needs:contract-review` review-chain bullets.
+ * Rulebook: `.claude/skills/pm-dispatch/references/contract-review.md` — the review of record and its `Served-tier:` line.
  */
 export const CONTRACT_REVIEW_TIER = 'claude-fable-5-1';
 

scripts/pm/dispatch-gates.mjs

@@ -297,35 +297,6 @@ for R in objectstack-ai/objectstack objectstack-ai/objectui objectstack-ai/cloud
   gh label create pm:epic             -R "$R" -c 5319e7 -d "Reserved by a dedicated epic PM, parent or sub-issue; other PMs never take it; never with pm:queue" 2>/dev/null || true
 done
 
-# needs:contract-review — the clause-② enqueue gate's re-review chain (SKILL.md
-# 入队与落地): a PR whose ACTUAL diff touches the contract surface — or whose
-# card's claim comment declares `Clause-②: yes` (the content limb, judged from
-# the card, path-independent) — but was dispatched below the contract-review
-# tier waits outside the queue under this label until the review sub-round
-# clears it. Named consumers: the enqueue gate and the triage sub-round's label
-# query. Main repo only — the contract surface (packages/spec) lives here. The
-# label names WHAT is reviewed, never a model; the tier's single source is
-# CONTRACT_REVIEW_TIER in scripts/pm/dispatch-gates.mjs. The -d below is kept
-# ≤100 chars (the hard cap — see the header). It did NOT match the live label
-# object: measured 2026-08-20, that object was grey with an EMPTY description,
-# so the claim this comment used to make had never been true — nothing had ever
-# checked it. `--reconcile` is what makes it true.
-#
-# ⚠️ WHAT THIS FILE DOES AND DOES NOT DO, because a card read it the other way
-# (#13922): everything here reconciles the label OBJECT — its colour and its
-# description. It says nothing about whether the label is HUNG correctly on any
-# pair, and it never did. Two other files answer that, and neither is this one:
-#   * `scripts/pm/check-half-states.mjs` H31 — the two carriers of one gate
-#     compared across the whole board, on the half-state patrol's schedule.
-#   * `scripts/pm/check-clause2-carriers.mjs` — the same comparison anchored on
-#     ONE pair (`--pair <pr>`, a pre-arm predicate), plus the two readings H31
-#     cannot make: the DECLARATION limb quoted three paragraphs up has no other
-#     reader in this repo, and a gate missing from BOTH carriers is a carrier
-#     comparison's silent case.
-# ⛔ None of the three writes the label. Hanging or clearing a review gate from
-# a script would be issuing the review verdict, which is 自查放行.
-gh label create needs:contract-review -R objectstack-ai/objectstack -c d93f0b -d "Clause-② enqueue gate: dispatched below contract-review tier — blocked until the review clears it" 2>/dev/null || true
-
 # needs:pack-smoke — the opt-in pre-merge pack smoke (#14214, derived from the
 # #14000 ruling; maintainer-approved 2026-09-01, director batch #23, 「同意」).
 # Its named consumer is a GATE rather than a query: the job guard in
@@ -408,8 +379,7 @@ gh label create repo:hotcrm   -R objectstack-ai/objectstack -c fbca04 -d "Seam c
 #
 # ⛔ Two acts this record deliberately does NOT perform. Stripping the label off
 # those 4 live carriers is a BOARD act for the PM at the landing window —
-# nothing in this file ever hangs or clears a label on a card (the
-# needs:contract-review block above states the same discipline). And the
+# nothing in this file ever hangs or clears a label on a card. And the
 # `repo:*` predicate in `scripts/pm/check-half-states.mjs` H14 stays LITERAL:
 # narrowing it is a predicate change ruled on #13992, never a side effect of a
 # vocabulary record. SKILL.md names no `repo:objectstack` (checked 2026-09-01),

scripts/pm/ensure-pm-labels.sh

@@ -1923,7 +1923,7 @@ export function claimKeyedLineRefusals(body) {
     rows.push({ key: 'Thread-read', line: keyLine(text, 'Thread-read'), why: `the value reads \`${offendingSpan(thread.value)}\`, which is neither one comment id nor \`none\` — H50 compares it for EQUALITY against one id, so no thread makes this match.` });
   const clause = readClause2Line(text);
   if (clause !== null && clause.kind !== 'declared')
-    rows.push({ key: 'Clause-②', line: clause.line, why: clause.kind === 'malformed' ? 'the value slot holds something `readClause2Line` cannot grade — the two spellings are the closed set.' : `\`readClause2Line\` reads this as a NEAR MISS (${clause.reason}), ⛔ not a declaration — \`check-clause2-carriers.mjs --pair\` answers exit 4 on it.` });
+    rows.push({ key: 'Clause-②', line: clause.line, why: clause.kind === 'malformed' ? 'the value slot holds something `readClause2Line` cannot grade — the two spellings are the closed set.' : `\`readClause2Line\` reads this as a NEAR MISS (${clause.reason}), ⛔ not a declaration — the changeset and ADR-0087 gates read it as absent.` });
   return rows;
 }