What the ruling assigned, and what appears to have no carrier
Ruling 5560224701 (director batch #60, 2026-09-06, maintainer verbatim 「同意」) chose a readable declaration-text snapshot for every @objectstack/spec export, and it named the mechanism that makes such a snapshot mean anything, verbatim:
A snapshot diff is not something a PR "accepts": it makes the PR a Clause-② carrier (needs:contract-review), and the contract-review seat reads the diff and rules minor/major. The existing review discipline is the guard against blind acceptance that a hash snapshot lacks.
and assigned the follow-up, also verbatim:
The contract-review reference (references/contract-review.md, skills seat) needs one input row naming the snapshot diff as a Clause-② signal; the director will raise it with the skills seat — not this card's scope.
The snapshot itself has now landed work behind it: PR #18971 (card #16045) creates packages/spec/api-surface-declarations/** (17 shards, 5,336 declaration blocks) and retires api-surface-signatures.json. That PR correctly did NOT wire the new artefact into any Clause-② detector, because the ruling put that row out of its scope by name.
Measured consequence: the guard is currently inert
Measured by the at-tier contract reviewer on PR #18971's head b33898f5d8d0dc8eff732eb8ac26452506b8622e (2026-09-18T0902Z), and recorded here because it is the ruling's own load-bearing mechanism:
PUBLISHED_SURFACES in scripts/pm/check-widening-tells.mjs is REGEN_ARTIFACTS.filter(row => row.check === 'check:api-surface') — a literal filter on the check name.- The new artefact's
REGEN_ARTIFACTS row carries check: 'check:api-surface-declarations', so it is off that surface by construction, not by omission.
⇒ Until the assigned row exists, a pull request that regenerates the shards beside a reshaped declaration and declares Clause-②: no raises no tell and passes check-clause2-carriers. The snapshot is readable, but nothing turns reading it into an obligation — which is precisely the "blind acceptance" the ruling chose text over a hash to prevent.
Why this is filed rather than asked in a comment
A cross-seat request is work, and work belongs on a card in the target lane's queue rather than in a comment that no sweep reads. The ruling says "the director will raise it"; this card exists so that the assignment has a carrier regardless of whether that hand-off happened.
⛔ This card does not prescribe the mechanism. Whether the signal should be the reference-doc input row alone (as the ruling says), or additionally something mechanised such as a PUBLISHED_SURFACES entry or a dedicated tell, is the skills seat's call — the domain:spec seat has no standing to design it, and the ruling did not delegate that choice here.
Dedup note
⚠️ I did not dedupe this, and my one instrument was weak — do not read it as an absence. I listed 49 open domain:skills cards and matched TITLES only against snapshot|contract-review|clause|api-surface; 6 titles matched and none was this assignment (they concern check-clause2-carriers defects, check-single-claim-paths, os-dev.md, and a separate charter conflict). Bodies were not read, closed cards were not read, and /search/* is 403 in this session (sessions are bound to their configured repositories), so a card with a differently-worded title would not have shown up. If this duplicates an existing card, close it as the duplicate and keep the older one.
Dedup words: contract-review.md, snapshot diff, Clause-② signal, PUBLISHED_SURFACES, api-surface-declarations
Provenance: filed by the domain:spec seat 3 (session_019srGWGCBBCBHqcDoRZpQRh) out of the at-tier contract review of PR #18971, 2026-09-18T0913Z. ⛔ No domain:* or priority set here — this seat is an execution seat and does not route or grade bare cards; the ruling quoted above names the owning seat, so the routing is the ruling's, not a judgement of mine.
Generated by Claude Code
What the ruling assigned, and what appears to have no carrier
Ruling
5560224701(director batch #60, 2026-09-06, maintainer verbatim 「同意」) chose a readable declaration-text snapshot for every@objectstack/specexport, and it named the mechanism that makes such a snapshot mean anything, verbatim:and assigned the follow-up, also verbatim:
The snapshot itself has now landed work behind it: PR #18971 (card #16045) creates
packages/spec/api-surface-declarations/**(17 shards, 5,336 declaration blocks) and retiresapi-surface-signatures.json. That PR correctly did NOT wire the new artefact into any Clause-② detector, because the ruling put that row out of its scope by name.Measured consequence: the guard is currently inert
Measured by the at-tier contract reviewer on PR #18971's head
b33898f5d8d0dc8eff732eb8ac26452506b8622e(2026-09-18T0902Z), and recorded here because it is the ruling's own load-bearing mechanism:PUBLISHED_SURFACESinscripts/pm/check-widening-tells.mjsisREGEN_ARTIFACTS.filter(row => row.check === 'check:api-surface')— a literal filter on the check name.REGEN_ARTIFACTSrow carriescheck: 'check:api-surface-declarations', so it is off that surface by construction, not by omission.⇒ Until the assigned row exists, a pull request that regenerates the shards beside a reshaped declaration and declares
Clause-②: noraises no tell and passescheck-clause2-carriers. The snapshot is readable, but nothing turns reading it into an obligation — which is precisely the "blind acceptance" the ruling chose text over a hash to prevent.Why this is filed rather than asked in a comment
A cross-seat request is work, and work belongs on a card in the target lane's queue rather than in a comment that no sweep reads. The ruling says "the director will raise it"; this card exists so that the assignment has a carrier regardless of whether that hand-off happened.
⛔ This card does not prescribe the mechanism. Whether the signal should be the reference-doc input row alone (as the ruling says), or additionally something mechanised such as a
PUBLISHED_SURFACESentry or a dedicated tell, is the skills seat's call — thedomain:specseat has no standing to design it, and the ruling did not delegate that choice here.Dedup note
domain:skillscards and matched TITLES only againstsnapshot|contract-review|clause|api-surface; 6 titles matched and none was this assignment (they concerncheck-clause2-carriersdefects,check-single-claim-paths,os-dev.md, and a separate charter conflict). Bodies were not read, closed cards were not read, and/search/*is 403 in this session (sessions are bound to their configured repositories), so a card with a differently-worded title would not have shown up. If this duplicates an existing card, close it as the duplicate and keep the older one.Dedup words:
contract-review.md,snapshot diff,Clause-② signal,PUBLISHED_SURFACES,api-surface-declarationsProvenance: filed by the
domain:specseat 3 (session_019srGWGCBBCBHqcDoRZpQRh) out of the at-tier contract review of PR #18971, 2026-09-18T0913Z. ⛔ Nodomain:*or priority set here — this seat is an execution seat and does not route or grade bare cards; the ruling quoted above names the owning seat, so the routing is the ruling's, not a judgement of mine.Generated by Claude Code