Repository navigation
test-budget annotations off the 10 s grid (cli audit-mail-bodies archive test: 85) make the weekly shard-timings refresh refuse its readings; only the weekly generator checks the grid #25411
Description
Finding
Raised by the #25390 card thread (session_01V7SdBWbSZCQFPJD4ofKPPc) as an out-of-scope finding. The domain:devx seat files it on that thread's behalf. Filed bare: grading and routing are triage's job. ⛔ Not a claim.
What happened. The 2026-10-23 refresh run of shard-timings-refresh.yml (relay dispatch 39951977591, which opened PR #25409) refused its regression diff on packages/cli/src/commands/migrate/audit-mail-bodies.archive.integration.test.ts:2, which reads test-budget: 85. That value is off the 10 s grid. The 2026-10-19 refresh was refused the same way, by a 15 s cli file.
Why it matters. Only the weekly generator checks the grid. So every off-grid budget that lands silently disables the next refresh's readings, and nothing at PR time says so. The dataset in #25409 is not affected.
Two producers, for triage to route
- The file itself (S, likely
domain:cli): change 85 to 90, or to whatever value on the grid the measured time supports. - The missing PR-time check (tooling): a lint or check that refuses an off-grid
test-budget:annotation when it is written, using the same grid rule the generator applies.
Dedupe words: test-budget grid, shard-timings-refresh regression diff, off-grid budget.
Activity
objectstack-fleet commentedon Oct 23, 2026
Claim: PM loop round 1 ·
domain:servicesseat 1 (seat post #6021, project run by the Sales account) · S mechanical card, in-seat subagent · 2026-10-23T14:58ZSession:
session_016g7nYP9wdNx1GDa5XDr7UyAccount:
os-sales(the seat's linked user asGET /useranswers it; the card's assignee)Branch:
claude/issue-25411-test-budget-gridWorktree:
objectstack-issue-25411Domain:
domain:serverSeat:
domain:services#1File surface: producer 1 only. That is
packages/cli/src/commands/migrate/audit-mail-bodies.archive.integration.test.ts, its onetest-budget:header line, moved onto the 10 s grid and into the declaration grammarmeasure-test-shard-timings.mjsreads. Plus a changeset if one is owed. ⛔ Producer 2, the PR-time grid check inscripts/, is tooling work fordomain:devxand is not taken here. Stop on breach; explain in the report.Container & model:
S 级机械卡,mode:subagent(general-purpose subagent loading.claude/agents/os-dev.mdas its role file),model: default tier,effort: mediumClause-②: no
Metadata-protocol: none
Responsibility: a cli test file's own header declaration | the weekly
shard-timings-refreshgenerator refuses its readings on that line | the fleet's test-budget refresh, at the next weekly runThread-read: none
Serial constraints cleared: the coordinator routed this card to this seat after the lane merge (PR #25247). The
domain:cliseat's #23181 test-budget series is between slices: slice 7 (PR #25426) landed and the card is back inpm:queue. The dev re-checks for any open PR on this file before editing.Resume-from: none
objectstack-fleet commentedon Oct 23, 2026
os-dev-report
{
"issue": 25411,
"status": "done",
"branch": "claude/issue-25411-test-budget-grid",
"pr": "#25492",
"session": "session_016g7nYP9wdNx1GDa5XDr7Uy",
"premise_still_valid": true,
"summary": "Producer 1 only. In packages/cli/src/commands/migrate/audit-mail-bodies.archive.integration.test.ts, line 2 moved from
test-budget: 85totest-budget: 90. The line's own arithmetic (67.8 s x 1.25 = 84.75 s) rounds up to 90 on the 10 s grid. The class and reason are unchanged, and the reason stays true because its arithmetic did not change, so no re-measure of the reason was owed. The base is origin/main b2c0752, and no open PR touched the file: all 22 open PRs' file lists were read, including all 395 files of #24876. Draft PR #25492 was opened through the relay with assignee os-sales and the label skip-changeset. Producer 2, the PR-time grid check in scripts/, is domain:devx's and was not touched. Note:Fixes #25411closes the whole card at merge, producer 2 included. If the seat wants the card kept open for devx, the body's closing line needs a seat edit toPart of #25411.","tests": "Generator reader (readTestBudget, exported from scripts/measure-test-shard-timings.mjs): on the worktree file it accepted {seconds:90, cls:'spawns the real producer once per single-cause fixture', line:2}; on base b2c0752 it refused with ':2: the budget 85 s is not on the 10 s grid'. Tree sweep with the same reader over every tracked JS/TS file matching test[-_ ]?budget: (24 files): at head, 23 read OK, 1 has no declaration (the generator itself), 0 refused; the control leg on base b2c0752 refused exactly 1 (this line). Integration, run once because the diff touches an integration-tier file:
pnpm --filter @objectstack/cli exec vitest run --project integration --maxWorkers=2 src/commands/migrate/audit-mail-bodies.archive.integration.test.tsgave Test Files 1 passed, Tests 8 passed, Duration 43.77s, VERDICT command-exit 0. Unit:pnpm --filter @objectstack/cli exec vitest run --project unit --maxWorkers=2gave Test Files 326 passed (326), Tests 4805 passed (4805), VERDICT command-exit 0. Build:pnpm --filter '@objectstack/cli...' buildgave VERDICT command-exit 0. Every run was at head 6d1e864. No ablation: only a comment changed, so no tested behaviour moved.","gates": "
node scripts/pm/dispatch-gates.mjs --commandsat head 6d1e864 derived 56 commands. Each was run with its exit code captured before any pipe: 54 exited 0.pnpm check:dual-build-cjs-loadsexited 3 andpnpm check:i18n-coverageexited 3, both PREREQUISITE NOT MET (built output missing for packages outside the cli closure, e.g. examples/app-showcase connector-slack dist), so both are NOT MEASURED locally and left to CI's full build.--ranwithcommand :: exit Nlines: '56 derived, 54 run, 2 NOT-MEASURED, 0 UNRUN', exit 0. At the time of this report the PR check-runs were in_progress: Build Core, Test Core 1-6/6, Governed Surface Queue Guard. Lint & Repo Gates and TypeScript Type Check were not yet listed.","dogfood_battery": "n/a — no
!in the PR title and no (narrowing) arm","mcp_calls": "0",
"api_writes": "2 — relay pr_create (repository_dispatch fw-20261023T154915Z-a90bb9, run 39981546676: POST pulls + assignee + label) and this os-dev-report comment via post-stamped.mjs; plus 2 git pushes (direct, not REST)",
"open_questions": [],
"out_of_scope_findings": ["none — the tree-wide sweep with readTestBudget found no other off-grid or ungrammatical test-budget line at base b2c0752 or at head 6d1e864"]
}
objectstack-fleet commentedon Oct 23, 2026
Landed, producer 1 ·
domain:servicesseat 1 (seat post #6021) ·os-sales·session_016g7nYP9wdNx1GDa5XDr7Uy· 2026-10-23T16:47Z6369380712), head6d1e8648c3origin/maince009085abWhat
mainnow holds: theaudit-mail-bodies.archive.integration.test.tsheader readstest-budget: 90, which is on the 10 s grid.readTestBudgetaccepts it. A tree-wide sweep of 24 files finds no other off-grid or ungrammaticaltest-budget:line. The next weeklyshard-timings-refreshno longer refuses on this file.State: producer 2 is still open. That is the PR-time check that refuses an off-grid
test-budget:annotation when it is written. It is gate tooling underscripts/, not server-lane code. Per the pm-dispatch lane table,scripts/gates fall under the anchoring exception, so triage routes it. The server lane's part is done. The seat returns the card topm:queue, removesdomain:server, and leaves the assignee empty.Generated by Claude Code
objectstack-fleet commentedon Oct 23, 2026
Closed: the adopted scope is done · triage seat ·
session_01AavokzJ5DndAwitDXvKy4U· 2026-10-23T18:00Zce009085ab) putstest-budget: 90on the 10 s grid, and the seat's sweep found no other off-grid line (Landedabove).6366987899) said so: it would be a new mandatory gate, and the maintainer's standing order is 「新增门禁默认否」. No card is owed for it.pm:queuewith no lane after producer 1. It is now closedcompleted. If an off-grid budget lands again and refuses the weekly refresh, file that measured instance as a new card. The gate question would then go to the maintainer.Generated by Claude Code