[rebuild of #19474] spec(lint): wire the inert runtime-create rules for action / hook / report / skill / email_template / mapping — six types, one edit (ruling #203/4 group A+C) #19542

⭐ REBUILD of card #19474, whose original is unreachable. Filed by the domain:spec seat 4 (session_01AmH9bKvGoLjiY86Q4Z3og2, seat post #18917) on 2026-09-21, under the same instruction that rebuilt #19384 as #19541: the original was filed by the banned os-sam account.

⛔ The original is not deleted and ⛔ nothing here overrules it. Measured at this act, exit codes captured before any pipe:

⇒ Why a rebuild was not optional here

This card has a live delivery: PR #19517, draft, head 72de2b946301aa59f624da807ae7fc383e82b81e, 6 files, +1018 / −11. The unreadable card blocks it mechanically, ⛔ not cosmetically:

PM_SWEEP_REPO=objectstack-ai/objectstack node scripts/pm/check-clause2-carriers.mjs --pair 19517   → exit 2
✗ UNJUDGED: card #19474's labels could not be read.

and

node scripts/pm/label-write.mjs --issue 19474 --add domain:spec --dry-run   → exit 5
① read — COULD NOT READ: GET …/issues/19474 -> HTTP 404

⇒ the clause-② enqueue gate can never be satisfied for #19517 while its card cannot be read, and this seat cannot hang or strip needs:contract-review on the card side of the ruled dual carrier. The PR could never land. That is the whole reason this card exists.

The delivery's state, carried over so nothing is re-derived

⚠️ Round 2 has NOT been judged. Two dispatches of the delta review both died on their first model request with HTTP 429 「You've reached your Fable limit」 (claude-fable-5-1 = CONTRACT_REVIEW_TIER), and the maintainer has since stated 「fable 没有了」. The review is ABSENT — ⛔ not FAIL, ⛔ not PASS — and PR #19517 stays draft, out of the queue, until the tier question is settled.

⚠️ The dev took the review's route (b) and held skill out, so group A lands as three types, not four. Whether that is inside the ruling's logic or a scope reduction its author should decide is the open question the absent round-2 review was dispatched to answer. ⛔ It is not settled, and ⛔ this seat has not settled it.

The hand-off card the hold-out produced is #19527.

The original card, reproduced verbatim below, ⛔ not rewritten

Filed by the domain:spec execution seat 3 (os-sam, session_01HnRAeVTLJevtQ5iCPX6JSm) at 2026-09-21T02:02Z, as the split ruling 5754204885 on #19275 explicitly instructs this seat to perform (batch #203 item 4, letter B, maintainer 「203 同意」). ⛔ Not a new finding, ⛔ not a re-grading: priority:p1 and domain:spec are inherited from #19275, whose split this is.

The policy, ruled once for all ten types: 「declared ⇒ honoured; not honourable ⇒ retired」.

Read before writing: ruling 5754204885 · measurement round 5750725047 (the specification) · triage's grouping 5750895260. Sibling ruling for app/position: #19370 (B, same policy).

Scope — six types, ONE card

Group A — action · hook · report · skill. The ruling: 「wire the existing candidate rules (runtimeTypes rows / TYPE_TO_STACK_KEY rows already present, inert only because the gate filters on runtimeTypes first)」. ⭐ One execution card for the four, ⛔ not four — the ruling says so in those words.

Group C — email_template · mapping. 「wire the ledger-driven rule the same way」, and the ruling folds group C into this card.

⛔ Fences, quoted from the ruling rather than paraphrased

• ⛔ No ledger-population work is dispatched for group C: 「the empty warn maps stay empty until a real property needs a row — ⛔ no ledger-population work is dispatched (zero pull), the wiring is the whole deliverable」.
• ⛔ Do not widen scope to the other four of the ten. doc / external_catalog retire under a separate card; translation / tool are read first under two more. Each has its own carrier.
• ⛔ Do not re-litigate letter B. The policy is ruled.

Acceptance

Per type: a real runtime write that the wired rule must refuse is refused, plus a lit control that a good write passes. ⛔ A surfaces / runtimeTypes field that merely changed is not the deliverable.

⚠️ The ruling's own NOT MEASURED, carried in unchanged

Still NOT MEASURED on any of the ten: whether a wired rule fires on a real write; the dispatch order names it as the report's first reading.

⇒ acceptance is behavioural. ⛔ A diff that changes a declaration without a test proving a real write is judged does not close this card.

查重词

allowRuntimeCreate · runtimeTypes · TYPE_TO_STACK_KEY · inert runtime-create declaration · declared but never honoured

Generated by Claude Code