Repository navigation
hourly full run: red on main (CI) #22968
Description
os-hourly-full-run-ci — machine-findable marker for this generated card. ⛔ Do not delete this line: it is how the hourly full run finds this card instead of filing a new one every hour.
hourly full run: red on main (CI)
Swept 2026-10-12T07:49:20.424Z · run log · commit 19805d8956f6b6728b9f4945da377f83c40c2cf0 · conclusion failure.
CI runs the FULL battery on main every hour (#16467). A push run on
main tests only the packages the merge touched, and a merge-queue build tests only what the group
contained — so this hourly run is the only thing that says whether the WHOLE tree is green, and
this card is the only channel that reports it. Nothing is blocked by it.
⛔ The remedy is never to narrow, skip or delete the failing check to make the hourly run green.
Fix what it names and let the next hourly run refresh this card; a card nobody reopens is closed by
the next green run being uneventful, not by editing this one.
What landed since the last green hourly run
Range: 51b43a8dc4 (run 38206660804, the last green CI schedule run) .. 19805d8956 — compare
da9c2bf2f2feat(adr-0087): the seventh disposition category, runtime-export-retired — retired exports and members, proven by the compiler at both revs (feat(adr-0087): the seventh disposition category, runtime-export-retired — retired exports and members, proven by the compiler at both revs #22940)e5687cb6b3feat(spec): an email field declares itself a correspondent address for inbound mail (FieldSchema.mailAssociation) (feat(spec): an email field declares itself a correspondent address for inbound mail (FieldSchema.mailAssociation) #22943)21914c3fb3fix(objectql)!: a multi-valued reference element is held to the single value's stored form (finding: a multi-valued reference field (lookup/userwithmultiple: true) admits an element of the wrong shape: REST stores an embedded record[{ id }]with 201, where the single-valued field refuses itinvalid_type(ADR-0104 D1) #22908) (fix(objectql)!: a multi-valued reference element is held to the single value's stored form (#22908) #22947)e23eccb164feat(spec,metadata-protocol)!: rename the type registry's allowOrgOverride to allowEnvironmentOverlay, on the registry and on GET /meta/types (feat(spec,metadata-protocol)!: rename the type registry's allowOrgOverride to allowEnvironmentOverlay, on the registry and on GET /meta/types #22874)1f312fdfa6docs(readme): link the essay, the compare page, the when-to-use page and the deck (docs(readme): link the essay, the compare page, the when-to-use page and the deck #22955)5f4e8a544cchore(docs,scripts): system-context page states no census totals; the census gate keeps the register (chore(docs,scripts): system-context page states no census totals; the census gate keeps the register #22950)39cca271c7docs(blog): four inline SVG figures for 'The Ontology Is the Software' (docs(blog): four inline SVG figures for 'The Ontology Is the Software' #22958)8d8f965374fix(objectql)!: a predicate rule's read through a master-detail FK asks the caller's read door for the header first — an unreadable header answers as a nonexistent one ([Decision] security(objectql): a validation rule that reads one hop through a master-detail FK still reads the header elevated, so an unreadable header's answer varies with its stored state after ruling A #22892) (fix(objectql)!: a predicate rule's read through a master-detail FK asks the caller's read door for the header first — an unreadable header answers as a nonexistent one (#22892) #22956)246edf5b39docs(releases): the versioning policy states the 17.x launch-window rule and the v18 prerelease regime (docs(releases): the versioning policy states the 17.x launch-window rule and the v18 prerelease regime #22963)117817ec28fix(runtime): on a kernel-resolver host, resolve a service on the request's own kernel first — record-bound POST /actions stops answering 404 RECORD_NOT_FOUND (fix(runtime): on a kernel-resolver host, resolve a service on the request's own kernel first — record-bound POST /actions stops answering 404 RECORD_NOT_FOUND #22945)19805d8956feat(create-objectstack): the blank starter ships a test story on @objectstack/verify; two test gates read the tsconfig-exclude payload boundary (feat(create-objectstack): the blank starter ships a test story on @objectstack/verify; two test gates read the tsconfig-exclude payload boundary #22960)
Filed by.github/workflows/scheduled-full-run-card.yml. Generated by Claude Code
Activity
objectstack-fleet commentedon Oct 12, 2026
Reading ·
domain:devxseat 1 ·os-warren·session_012yK1ddAdAqfzwigRHZq3r2· 2026-10-12T06:37Z. ⛔ Not a claim, ⛔ no label change.21914c3fb3) has one failing step,Test Core→ "Check the per-package suite-duration ceilings". One package is over:@objectstack/cloud-connection: 167.04s, over its 163.89s ceiling. Slowest file: src/marketplace-install-local-seed-lookup.test.ts 8.56s.It was read throughscripts/pm/ci-failure.mjs --run 38210849670(check-run annotations).1bd612bd6; closed after the 05:08Z run read 117.99 s). None of the three commits in this card's range touchespackages/cloud-connection.Test Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949. Per the seat's second addendum there, this hourly reading means C2 alone would not keep this card closed. The recommendation now adds a ruled raise forcloud-connectionalone. No second decision is filed from here.Test Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949's ruling and closes on the first uneventful green hourly run after the ruled change lands. A further red on the same package is a sighting for ci: merge-queueTest Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949, not a new reading here.Generated by Claude Code
objectstack-fleet commentedon Oct 12, 2026
Reading and close ·
domain:devxseat 1 ·os-warren·session_012yK1ddAdAqfzwigRHZq3r2· 2026-10-12T09:04Z. Closednot_planned, following the #22904 and #22946 precedent: the next hourly run is green, and the decision behind this signature has landed.scheduleruns after this card was filed:38217670691(19805d8956, 07:25Z): red on the ceiling step, but on@objectstack/knowledge-ragflow(5.14 s against 3.40 s), notcloud-connection. So there is no second consecutive red on one package, which ci: merge-queueTest Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949's ruling counts as a reading. The package's 3.40 s ceiling is now under the 20 s floor, so it is never red again.38222448090(712361b7a8, 08:15Z):success. InTest Core, "Check the per-package suite-duration ceilings" issuccess, and all six shards' "Check this shard's timing drift" steps aresuccess.main: PR ci: Test Core's two timing gates exit-grade onschedulealone (C2), with a 20s ceiling floor #22973 merged aseede467f3(08:33Z), with C2 for both timing steps plus the 20 s floor (ci: merge-queueTest Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949, ruling6119186511, landing record on that card). From the first hourly run on or aftereede467f3, a red on either timing step is aschedulered by design. Two consecutive reds on one package are a reading, posted on ci: merge-queueTest Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949 under its ruling.Generated by Claude Code