dogfood: main's Dogfood Regression Gate (3/3) is red since #25956 — the authz blind-spot census still pins handleI18nRequest at 3 in packages/runtime/src/domains/i18n.ts; #25956 made it 4

Filing gate ③: maintainer-directed task. The maintainer, 2026-10-26 08:01Z: 「main 上的 Dogfood Regression Gate (3/3) 时红时绿，跟合仓无关，还没有 issue 跟踪」. Filed by the devx seat (seat post #6023, session session_01LKHDniRQSWjRJyDtacwpEd). The claim follows separately.

Business reason: Dogfood Regression Gate (3/3) is red on every main run that runs it, starting at #25956 (06:42Z today). The release gate (full-battery) reads those runs. A red full battery at the version commit refuses a cut, and every red push run also re-files #25438.

Reader who acts: domain:devx. packages/qa/dogfood is the QA harness, and the server lane is not seated today. One small PR.

It looks intermittent, but it is a deterministic red

The red comes from one assertion. It fails the same way on every run that runs shard 3/3:

FAIL [isolated] test/authz-probe-blind-spot.test.ts > authz probe blind-spot census (#13260)
  > packages/runtime/src/domains/i18n.ts — every positive control is still present in THAT file
AssertionError: control "handleI18nRequest" moved in packages/runtime/src/domains/i18n.ts: expected 4 to be 3

Cause: 69fa23d60b (fix(runtime): GET /i18n/translations/:locale carries an ETag and Cache-Control: private, no-cache, and a matching If-None-Match is answered 304 (#25785) #25956, Part of rest/i18n: GET /i18n/translations/:locale serves the whole locale bundle (288–361 KB raw, 80–88 KB compressed) on every console open with no ETag or Cache-Control, and carries Studio's metadataForms (46.8 KB) into the boot read #25785, server card thread). It added a 4th occurrence of handleI18nRequest to packages/runtime/src/domains/i18n.ts. On origin/main the file holds 4 occurrences (lines 20 and 73 are docblock {@link handleI18nRequest}, 150 is the call, 155 is the declaration); at 37617109 it held 3. PROBE_FILE_CENSUS still records 3.
Why it looks intermittent: a push run on main is affected-only (ci: push runs on main use the affected set; an hourly scheduled run keeps the full list and files a card on red (maintainer-directed) #16467), so a merge that does not touch the dogfood closure skips the Dogfood Regression Gate (N/3) legs. The aggregate is then green, for example 0c37228's run 40282257384. pull_request and merge_group never run the dogfood legs at all (dispatch-gates lists Dogfood Regression Gate as "scheduled BY YOUR PATHS" for a route-ledger.ts change, but its if: skips pull_request and merge_group, so neither PR CI nor the derived commands run the dogfood pins (main red after #24009) #24073), which is how fix(runtime): GET /i18n/translations/:locale carries an ETag and Cache-Control: private, no-cache, and a matching If-None-Match is answered 304 (#25785) #25956 merged green.
Reproduced locally on origin/main cdd4eea938: npx vitest run test/authz-probe-blind-spot.test.ts in packages/qa/dogfood → 1 failed | 46 passed, the same assertion.

Runs on `main` (CI workflow), newest first

run	event	head	`Dogfood Regression Gate (3/3)`
40284424233	push	`cdd4eea`	❌ failure (job 121444813097)
40283575048	push	`963e504`	❌ failure (job 121442160635)
40282257384	push	`0c37228`	skipped (affected-only), aggregate ✅
40281628332	push	`38050ce`	❌ failure (job 121436163074)
40281437744	push	`69fa23d` (#25956)	❌ failure (job 121435563374), first red
40281225169	push	`3761710`	✅ success (job 121434899004), last green before the cause
40280560677	push	`d5f4c2a`	✅ success
40277513449	schedule	`9ca36b1`	✅ success

Before 69fa23d, the 3/3 leg was green on all 56 main runs from 2026-10-25 05:57Z to 06:40Z today that ran it. The red CI conclusions in that window came from other jobs (#25438 tracks those).

Wanted

The census is a measurement pin. Its own header says that whoever moves a count re-reads it, so don't just bump the number. Read the 4th occurrence. If it is only the new docblock {@link} and the probe's reach is unchanged (the population and reachable rows for i18n.ts still pass today), record 4 and say so in the PR body.
If the new code added a handler or route that the probe should see, stop and report it. That is a census row change, not a count bump.
Check whether any other control or row in authz-probe-blind-spot.census.ts moved with fix(runtime): GET /i18n/translations/:locale carries an ETag and Cache-Control: private, no-cache, and a matching If-None-Match is answered 304 (#25785) #25956 or with later merges. Run the whole packages/qa/dogfood shard 3/3 locally (or the census + authz-conformance files) before you push.
Acceptance: Dogfood Regression Gate (3/3) is green on the first main push run after landing that runs it, and on the next scheduled run. Report the run ids on this card.

Not in scope

Making PR or merge-group CI run the dogfood legs. That was ruled on in dispatch-gates lists Dogfood Regression Gate as "scheduled BY YOUR PATHS" for a route-ledger.ts change, but its if: skips pull_request and merge_group, so neither PR CI nor the derived commands run the dogfood pins (main red after #24009) #24073 / [Decision] The required Test Core and Dogfood Regression Gate go red on superseded runs (139 in 51 h, 0 real catches): keep #16157's ruled reading, or decide by the cancel cause the platform records? #22653, and is a separate decision.
The ui-merge Test UI red (check-pre-install-import-graph, scripts/objectstack-pin.mjs), which is ci(ui-merge): the pre-install import graph reads the repository root's workflows; objectui's prompt files move to the root .github/prompts/ #25957 / tooling(ui-merge): governance in-repo — liveness evidence and spec tests stop citing objectui by commit, one sdui-parser, AGENTS.md / pm scripts drop objectui#, ui/AGENTS.md nests, 18 governance workflows merged or retired (parent #25677, 4b/6) #25712.

Clause-②: no
Metadata-protocol: none

Generated by Claude Code

Claim: PM loop · 2026-10-26T08:11Z
Session: session_01H98bQgxPLVbpyR6Zsu6SUW (card thread, dispatched by seat session session_01LKHDniRQSWjRJyDtacwpEd)
Account: os-elon-musk
Branch: claude/project-thread-vi44lw (the harness-pinned branch for the card thread; it names no issue, so this claim is the identity record)
Worktree: objectstack-issue-25988
Domain: domain:devx
Seat: domain:devx#1 (seat post #6023)
File surface: packages/qa/dogfood/test/authz-probe-blind-spot.census.ts (+ its test only if a row changes) (stop on breach; explain in the report)
Container & model: S (p1), mode:thread, model: default tier, effort: xhigh
Clause-②: no
Metadata-protocol: none
Responsibility: n/a — measurement pin re-read
Thread-read: none
Serial constraints cleared: none in flight on packages/qa/dogfood

Dispatched under the maintainer's 「同意」 to the director's ruling batch 1, item 5A (devx dispatches p1–p3). Dispatch order: /mnt/project-files/pm-devx-seat/dispatch-25988.md.

os-dev-report

{
  "issue": 25988,
  "status": "done",
  "branch": "claude/project-thread-vi44lw",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/26002",
  "session": "session_01H98bQgxPLVbpyR6Zsu6SUW — card thread (dispatched by seat session session_01LKHDniRQSWjRJyDtacwpEd)",
  "premise_still_valid": true,
  "summary": "Premise holds: at base 05f7758f the census test reads 1 failed | 46 passed on exactly the i18n.ts handleI18nRequest control (4 vs 3). The 4th occurrence is the {@link handleI18nRequest} in TRANSLATIONS_CACHE_CONTROL's docblock (i18n.ts:73), added by 69fa23d6 with three module-private helpers and no handler, DomainRoute prefix or shouldDenyAnonymous( call, so population / reach / gate control stay 1 / 1 / 1. The row's control is recorded at 4 and its comment names the 4th occurrence; nothing else in the census moved (47/47 green with the change). One file, +12 / -4.",
  "tests": "head ed1c5b39, all under os-verify-lock: closure build `pnpm turbo run build --filter=@objectstack/dogfood --concurrency=2` → Tasks 67/67, VERDICT command-exit 0. Shard `OS_TEST_SHARD=3/3 node scripts/run-with-stall-guard.mjs ... -- pnpm turbo run test --filter=@objectstack/dogfood` → Test Files 97 passed | 2 skipped (99), Tests 990 passed | 14 skipped (1004), VERDICT command-exit 0; check-test-completeness OK 1004/1004. authz-conformance.test.ts + authz-probe-blind-spot.test.ts → 2/2 files, 108/108 tests. Census alone at base 05f7758f before the change → 1 failed | 46 passed (the issue's assertion). `pnpm --filter @objectstack/dogfood typecheck` → VERDICT command-exit 0. No ablation: the diff changes a recorded measurement, not tested behaviour.",
  "gates": "dispatch-gates --commands at ed1c5b39: 52 families on 1 path; 51 exit 0; pnpm check:dual-build-cjs-loads exit 3 (PREREQUISITE NOT MET: needs every package's dist, only the dogfood closure was built) = NOT MEASURED, CI Lint owns it. `dispatch-gates --ran`: 52 derived, 51 run, 1 NOT-MEASURED, 0 UNRUN, exit 0. eslint --no-inline-config on the changed file exit 0. check-governed-merges --branch HEAD: NOT governed, 16 changed lines. PR check-runs: in_progress at report time.",
  "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm; shard 3/3 was run anyway per the dispatch order: ed1c5b39 · VERDICT command-exit 0 · Test Files 97/99 (2 skipped) · Tests 990/1004 (14 skipped) · no failures",
  "mcp_calls": "3 — mcp__github__issue_read x2 (get, get_comments on #25988), mcp__github__pull_request_read x1 (get #26002); no write tool",
  "api_writes": "2 relay strokes — (1) pr_create #26002 (POST /pulls + assignees os-elon-musk + labels skip-changeset), run 40293605219; (2) pr_request_reviewers #26002 (os-elon-musk) + this comment on #25988",
  "open_questions": [],
  "out_of_scope_findings": []
}

Generated by Claude Code

Seat ACCEPT on PR #26002 at head ed1c5b3 (re-read pin: the 4th occurrence is #25956's docblock link; population/reach/gate 1/1/1 unchanged). Lands when CI is green.

Generated by Claude Code

Acceptance readings for #25988, after #26002 landed (merge a3e5bfa6, 2026-10-26 09:11Z). Both are green.

reading	CI run	head	`Dogfood Regression Gate (3/3)`
first `main` push run that runs the leg	40297613977 (push)	`a3e5bfa6` (#26002 itself)	job 121487638380, success, 09:18:22Z. All 23 jobs in the run were success or skipped.
next scheduled run	40330891455 (schedule, the 12:00Z slot, created 13:28:36Z)	`95294be3`	job 121599940412, success, 13:34:34Z. Dogfood (1/3), (2/3) and the `Dogfood Regression Gate` aggregate were also success. When this was read, 36 of the 37 jobs had finished, all success, and `Test Core (4/8)` was still running.

The census assertion that reddened every 3/3 leg since 69fa23d6 is gone on both runs. The card's acceptance is met, and it goes back to the devx seat for closing.

Generated by Claude Code

Navigation Menu

Repository navigation

dogfood: main's Dogfood Regression Gate (3/3) is red since #25956 — the authz blind-spot census still pins `handleI18nRequest` at 3 in packages/runtime/src/domains/i18n.ts; #25956 made it 4 #25988

It looks intermittent, but it is a deterministic red

Runs on `main` (CI workflow), newest first

Wanted

Not in scope

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

dogfood: main's Dogfood Regression Gate (3/3) is red since #25956 — the authz blind-spot census still pins handleI18nRequest at 3 in packages/runtime/src/domains/i18n.ts; #25956 made it 4 #25988

Description

It looks intermittent, but it is a deterministic red

Runs on main (CI workflow), newest first

Wanted

Not in scope

Activity

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

dogfood: main's Dogfood Regression Gate (3/3) is red since #25956 — the authz blind-spot census still pins `handleI18nRequest` at 3 in packages/runtime/src/domains/i18n.ts; #25956 made it 4 #25988

Runs on `main` (CI workflow), newest first