gate: enforce the quarantine deletion ratchet — nothing ran it, and it could not fail (#3167)

## A policy with nothing enforcing it

AGENTS.md states the deletion ratchet plainly:

> A quarantined test is **DELETED after 14 days** (`quarantinedAt` + 2
weeks) unless rescued.

Nothing enforced it, and it failed in two independent ways:

1. **`check:quarantine-ledger` omitted `--strict`.** The script only
exits non-zero with that flag (`check-quarantine-ledger.mjs:197`:
`return args.strict && (summary.expired > 0 || summary.near > 0) ? 1 :
0`). Without it, it is a report that always exits 0.
2. **No workflow ran it.** I audited all 12 `check:*` scripts against
`pr-checks.yml` and `full-suite.yml`: this is the only one appearing in
neither.

Either alone would have made it toothless. Together, a quarantined test
could sit past its deletion date indefinitely with every gate green.

## This exact shape is already documented in the file I edited

The comment above the lifecycle-column ratchet in `pr-checks.yml` says:

> `pnpm census:lifecycle-columns` (no `--strict`) and nothing ran it, so
three PRs lowered counts without re-recording and left allowances the
deleted guards could return through while this gate stayed green.

Script supports enforcement → package script omits the flag → no
workflow runs it. Same three steps, different ratchet. That precedent is
why I went looking.

## What changed

- `check:quarantine-ledger` now passes `--strict`
- wired into `pr-checks.yml` beside its siblings

It fires **5 days before** the deadline, not after, so the response is
still delete-or-rescue rather than an overdue entry. That window is the
script author's design; I did not invent it.

## Mutation evidence

Against a temp ledger, real one restored after:

| Ledger state | Result |
|---|---|
| today (1 entry, 13 days remaining) | exit 0 |
| entry inside the 5-day near window | **exit 1** |
| entry 6 days past deadline | **exit 1**, reports `EXPIRED (6 days
overdue)` |
| real ledger restored | exit 0 |

Without `--strict` all four exit 0 — which is the state on `main`.

## Honest note on what this will do

This is a **deadline ratchet**: it fires on a timer by design. The
current entry (`useTasks-hydration-freshness.test.ts`, deadline
2026-08-13) will trip it on **2026-08-08** unless someone deletes or
rescues it first. That is the intended behaviour and the whole point —
AGENTS.md is explicit that rescue "requires evidence the test catches
real regressions plus a root-cause fix — not stabilization passes." A
gate that never fires enforces nothing.

## Census before / after

```
before:  COLUMN guards (the backlog):   13
after:   COLUMN guards (the backlog):   13
```

Unchanged — this touches no lifecycle guard. It is gate wiring.

## Verification

`test:gate` exit 0 · `check:quarantine-ledger --strict` exit 0 on the
real ledger · both failure arms mutation-verified · ledger file restored
byte-for-byte.
This commit is contained in:
gsxdsm
2026-07-31 08:17:00 -07:00
committed by GitHub
parent 7d9d097acf
commit bc37185026
2 changed files with 8 additions and 1 deletions

View File

@@ -66,6 +66,13 @@ jobs:
run: pnpm check:lane-wiring
- name: Plugin interop declarations match the dashboard API
run: pnpm check:plugin-interop-drift
# AGENTS.md states the deletion ratchet as policy — "A quarantined test is DELETED after 14
# days (quarantinedAt + 2 weeks) unless rescued" — and nothing enforced it. Same shape as the
# lifecycle-column note above: the script supports `--strict`, the package script omitted it,
# and no workflow ran it, so it could not fail however long an entry sat there. It fires 5 days
# BEFORE the deadline so the choice is still delete-or-rescue rather than an overdue entry.
- name: Quarantine deletion ratchet
run: pnpm check:quarantine-ledger
typecheck:
name: Typecheck

View File

@@ -27,7 +27,7 @@
"check:lane-wiring": "node scripts/check-lane-wiring.mjs",
"check:plugin-interop-drift": "node scripts/check-plugin-interop-drift.mjs",
"census:lifecycle-columns": "node scripts/lifecycle-column-census.mjs",
"check:quarantine-ledger": "node scripts/check-quarantine-ledger.mjs",
"check:quarantine-ledger": "node scripts/check-quarantine-ledger.mjs --strict",
"check:mock-completeness": "node scripts/check-mock-completeness.mjs",
"test:gate": "node scripts/check-no-nohup.mjs && node scripts/check-no-cwd-relative-dashboard-test-reads.mjs && node scripts/check-no-kill-4040.mjs && node scripts/check-no-getdatabase.mjs && node scripts/check-capacity-pool-id.mjs && node scripts/check-no-node-only-core-imports-in-dashboard.mjs && node scripts/check-pi-versions-pinned.mjs && node scripts/check-no-test-timeout-appeasement.mjs && node scripts/check-changeset-format.mjs && node scripts/check-mock-completeness.mjs && node scripts/check-inert-sync-lane-conversions.mjs && sh -c 'pnpm --filter @fusion/engine test:core & engine_pid=$!; pnpm --filter @fusion/core test:pg-gate & pg_pid=$!; pnpm --filter @fusion/core test:unit-gate & unit_pid=$!; status=0; wait $engine_pid || status=1; wait $pg_pid || status=1; wait $unit_pid || status=1; exit $status' && pnpm --filter @runfusion/fusion test:ci-shape",
"smoke:boot": "node scripts/boot-smoke.mjs",