[Decision] Merge queue: stop the remaining test shards once one shard is red (merge_group only), which narrows ruling 6139629628 A1 inside the queue #31517

Activity

1. objectstack-fleet

   added

   toolingFix lands in tooling, not a product package — enters pm:queue only with Unblocks: or a named surfaceFix lands in tooling, not a product package — enters pm:queue only with Unblocks: or a named surface

   priority:p1High: required for production / M2High: required for production / M2

   needs-user-decisionBlocked on a maintainer decision — do not dispatch; decision cards only, never on a PRBlocked on a maintainer decision — do not dispatch; decision cards only, never on a PR

   domain:devxDomain lane — seat card indexed by label:pm:seatDomain lane — seat card indexed by label:pm:seat

   area:devpathThe road — create, dev, verify, publish/install, connect an agent, iterateThe road — create, dev, verify, publish/install, connect an agent, iterate

   on Nov 8, 2026
2. objectstack-fleet commented on Nov 9, 2026

   

   objectstack-fleetBot

   on Nov 9, 2026

   ContributorAuthor

   More actions

   triage#1 · os-zhuang · session_01AavokzJ5DndAwitDXvKy4U

   Ruling: #31517 · letter A · maintainer 「31517 … 同意」 2026-11-09T01:23Z

   The maintainer asked triage for an explanation (「具体解释 #31517…」), and then answered 「31517 31115 29471 31505 同意」 to triage's recommendation, A.

   • What is ruled: queue-only fail-fast. fail-fast: ${{ github.event_name == 'merge_group' }} goes on test, test-ui and dogfood. PR, push and schedule runs stay complete, so [Decision] 门禁过严拖慢发版:一轮 CI 暴露全部红、成本型棘轮加容差带、门禁瘦身审计 #23177 A1's 「一轮 CI 暴露全部红」 still holds where the author reads it.
   • Binding constraint: use the matrix's own fail-fast only, never a run-level cancel (CI: Dogfood Regression Gate 把 cancelled 当失败 —— 每次连续推送都产生一条假红 #3668: a cancelled matrix with zero attestations passes its aggregate).
   • Execution: domain:devx, one Tier H PR on ci.yml, plus self-test fixtures if they pin fail-fast: false. Read the first week of queue data after landing.
   • needs-user-decision removed, pm:queue.

   ---

   Generated by Claude Code
3. objectstack-fleet

   added

   pm:queueReady for the PM dispatch loopReady for the PM dispatch loop

   pm:dispatchedDispatched to a dev agent by /pm-dispatchDispatched to a dev agent by /pm-dispatch

   and removed

   needs-user-decisionBlocked on a maintainer decision — do not dispatch; decision cards only, never on a PRBlocked on a maintainer decision — do not dispatch; decision cards only, never on a PR

   pm:queueReady for the PM dispatch loopReady for the PM dispatch loop

   on Nov 9, 2026
4. objectstack-fleet

   assigned

   zhuangjianguo

   on Nov 9, 2026
5. objectstack-fleet commented on Nov 9, 2026

   

   objectstack-fleetBot

   on Nov 9, 2026

   ContributorAuthor

   More actions

   domain:devx#2 · zhuangjianguo · session_01QsR2jTo666XXx9e1Gm21C6

   Claim: PM loop round 9 · 2026-11-09T01:27Z
   Session: session_01QsR2jSkaG5iii6WC7w4qSX (card thread, dev) · seat session_01QsR2jTo666XXx9e1Gm21C6
   Account: zhuangjianguo
   Branch: claude/project-thread-c9oqlr
   Worktree: objectstack-issue-31517
   Domain: domain:devx
   Seat: domain:devx#2
   File surface: .github/workflows/ci.yml (strategy.fail-fast on test, test-ui, dogfood, and the A1 comments), and the self-tests and fixtures that read them (scripts/check-shard-attestation.mjs, check-aggregator-roster.mjs, partition-test-shards.mjs, and any other a grep finds). Tier H, hand-merged. ⛔ No run-level cancel, no concurrency change, no other workflow, no ui/.github/**.
   Container & model: S, mode:thread, model: default tier (no path-derived mandate), effort: high
   Clause-②: no
   Metadata-protocol: none
   Responsibility: merge-queue test matrices | a red shard waits for the slowest sibling (Test Core 4/8 median 19.0 min) before the aggregate reds | every queue entry behind a red one
   Thread-read: 6730013883
   Premise: holds; ruling A recorded in 6730013883, and fail-fast: false stands on the three matrices on main
   Serial constraints cleared: PR #31611 (#31519) and PR #31644 (#22034) landed in ci.yml. The dev confirms that a fail-fast-cancelled queue run is never read as green by --queue-verdict or the borrowed Test UI decision.

   Scope: one PR as ruled. The PR says Fixes #31517. Brief: /mnt/project-files/issue-31517/brief.md. Summons: ruling 6730013883 and mailbox notice 6730034988.
6. objectstack-fleet commented on Nov 9, 2026

   

   objectstack-fleetBot

   on Nov 9, 2026

   ContributorAuthor

   More actions

   domain:devx#2 · zhuangjianguo · session_01QsR2jSkaG5iii6WC7w4qSX
   os-dev-report

   {
     "issue": 31517,
     "status": "done",
     "branch": "claude/project-thread-c9oqlr",
     "pr": "https://github.com/objectstack-ai/objectstack/pull/31753",
     "head": "0a383d677bd8bd31bdb92f4778c644ced0bce6f9",
     "session": "session_01QsR2jSkaG5iii6WC7w4qSX",
     "premise_still_valid": true,
     "summary": "Ruling A executed: strategy.fail-fast is ${{ github.event_name == 'merge_group' }} on test, test-ui and dogfood in ci.yml; pull_request, push, schedule and workflow_dispatch keep every shard running. The A1 comments now say the queue is the exception across shards and why. check-shard-attestation's self-test gains an 8-case battery for the fail-fast shape (floor 25 -> 26) using the measured cancel annotation of run 30271824408 (check run 89995981782: 'The strategy configuration was canceled because \"matrix._1\" failed'); ci-shard-matrix's fixture spells the new expression. Confirmed without change: queueRunVerdict covers only a success queue run and testUiFullGreen counts only success shards, so a fail-fast queue run is never borrowed by a push.",
     "files_changed": [".github/workflows/ci.yml", "scripts/check-shard-attestation.mjs", "scripts/ci-shard-matrix.mjs"],
     "line_budget": "97 changed lines (+90 / -7), 3 files, under the 3000 threshold",
     "tests": "At 0a383d67: check-shard-attestation --self-test exit 0 (350 assertions), live exit 0; ci-shard-matrix --self-test exit 0 (27 cases); check-aggregator-roster --self-test and live exit 0; partition-test-shards --self-test exit 0; scheduled-full-run --self-test exit 0; ui/scripts/__tests__ select-affected-tests-wiring, select-affected-tests, test-file-timings: 3 files, 95 tests passed. Mutation probes on the committed head 6903a73d: classifier reading the fail-fast text as a supersession -> exit 1 (2 #31517 cases red); judge routing failure into the cancelled branch -> exit 1 (1 #31517 case red); both restored, blob hash equal to HEAD.",
     "gates": "dispatch-gates --commands at 0a383d67: 63 derived, 63 run, all exit 0; --ran: '63 derived famil(ies) accounted for — 63 run, 0 NOT-MEASURED'. First run at 6903a73d: 59 exit 0, 4 exit 3 (dts-closure, dual-build-cjs-loads, lean-entry-closure, sourcemap-no-sources-content: no dist); type-check-debt built the workspace there, so the final run reads a dist built at 6903a73d. Body gates on the PR body: check-changeset-no-major, check-adr-0087-registration, check-closing-keyword-parity, check-partof-closing-keyword, all exit 0. PR check-runs: in_progress at report time.",
     "dogfood_battery": "n/a — no ! in the PR title and no (narrowing) arm",
     "merge_tree": "git merge-tree HEAD origin/main exit 0 before merging 90cea476; no parallel fail-fast change on main in the touched files",
     "deviations": "none",
     "mcp_calls": "6 — issue_read x3 (get, get_comments x2), actions_list list_workflow_jobs x1, get_check_run x1, pull_request_read get x1; all reads, no MCP write",
     "api_writes": "2 relay writes — pr_create (fleet-write run 42551655351, read back identical) and this post-stamped comment; plus git push x2; 2 REST reads of check-run annotations (89995981782, 89995981757)",
     "open_questions": [],
     "security_findings": "none",
     "out_of_scope_findings": ["carrier: none · noted, not filed: a matrix whose every shard is killed by its own timeout before any attests reads cancelled with zero attestations and passes by #3668 (ruled not extended on #22653); fail-fast does not widen it"],
     "owed_after_landing": "devx seat: 7-day reading on #31517 of merge_group queue time from the first red shard to the aggregate red, against the 7 days before"
   }
7. objectstack-fleet commented on Nov 9, 2026

   

   objectstack-fleetBot

   on Nov 9, 2026

   ContributorAuthor

   More actions

   domain:devx#2 · zhuangjianguo · session_01QsR2jTo666XXx9e1Gm21C6

   PR #31753 is ACCEPTed and waits on the maintainer's hand merge (Tier H): head 0a383d677bd8bd31bdb92f4778c644ced0bce6f9, ACCEPT 6731174723, contract review PASS 6731165523 (nits only). The queue's test, dogfood and test-ui matrices become fail-fast on merge_group only, by the matrix's own fail-fast, as ruled. When it lands, this seat writes the landing note and owes the 7-day queue reading, including the count of queue runs whose aggregate passed on a cancelled zero-attestation matrix (expected 0).
8. os-zhuang

   closed this as completedin #31753on Nov 9, 2026
9. objectstack-fleet commented on Nov 9, 2026

   

   objectstack-fleetBot

   on Nov 9, 2026

   ContributorAuthor

   More actions

   domain:devx#2 · zhuangjianguo · session_01QsR2jTo666XXx9e1Gm21C6

   Landed · PR #31753 was merged by the maintainer (os-zhuang) through the merge queue as e253f3ab09717234fedae8e4d7a193cd1a12b0da. It is Tier H. The seat posted ACCEPT 6731174723 on head 0a383d67, and the contract review PASS is 6731165523, nits only. From this commit on, the test, dogfood and test-ui matrices are fail-fast on merge_group only, through each matrix's own fail-fast, as ruling A (6730013883) says. Every other event still runs every shard. A fail-fast queue run concludes failure, so a push never borrows it.

   Owed by this seat on this card after 2026-11-16T03:45Z, the 7-day after-reading:

   1. On red queue runs, the queue wall from the first red shard to the aggregate red, and the runner minutes, compared with the pre-change baseline (a red shard waited for the slowest sibling, Test Core 4/8 median 19.0 min).
   2. The count of merge_group runs whose Test Core or Dogfood Regression Gate passed on a cancelled matrix with zero attestations (expected 0, from contract-review nit N3).

   The record's nits N1 and N2 (comment wording and wrap in ci.yml) ride the next ci.yml touch.
10. objectstack-fleet

    removed

    pm:dispatchedDispatched to a dev agent by /pm-dispatchDispatched to a dev agent by /pm-dispatch

    on Nov 9, 2026