docs(issues): file the merged-PR re-fire defect in the PR mergeability check

[BigSimmo]

Summary

• Files one immutable inbox request (docs/outstanding-issues-inbox/8dfc4aaf-…json, P2, issue) recording a confirmed defect: the mergeability job in .github/workflows/pr-mergeability.yml has no filter on pull_request.state / pull_request.merged, while its pull_request_target trigger includes edited — a type GitHub fires on merged and closed pull requests. A merged PR reports mergeable_state: "unknown" permanently, which is exactly the retry branch of classifyMergeability() in scripts/pr-mergeability.mjs, so the required check fails deterministically after its five attempts. Confirmed on PR docs(issues): reconcile 38 queued ledger requests into the canonical ledger #2229: workflow run 32489311892, job 96792972554, failed 97 seconds after the merge, triggered by a bot description edit.
• The request also records a correction to the original report. PR required and Static PR checks did not re-fire post-merge — ci.yml's pull_request types do not include edited, every ci.yml run on claude/ledger-reconcile-0821b is pre-merge, and the last one (32489042816, head 754cbfcb) was green three seconds before the merge. The appearance of post-merge failure came from ci-triage.yml editing one comment in place: comment 5369760686 was created at 12:23:46Z and last updated at 13:44:36Z, both pre-merge, but still named those two jobs on the page after the merge.
• The proposed fix and the reason it is deliberately not applied here are recorded in the request rather than implemented, because .github/workflows/** is operational-risk scope and needs its own PR with a workflow contract test.

No workflow file, script, or canonical ledger row is changed by this PR. The canonical docs/outstanding-issues.md is untouched; a serialized reconcile branch applies the request later.

Verification

• npm run verify:pr-local — recognised low-risk documentation scope, exit code 0 captured directly rather than through a pipe. Decisive lines: Ledger inbox check passed: 6 pending request(s), 485 applied., Outstanding-issues guard passed: 423 rows (67 open, 356 archived), unique display and durable ids, collision-free allocation enabled, deprecated next-id marker ignored, no merge driver, no ids deleted from base 3c94d555357d., and Ledger write discipline passed for 3c94d555357d..HEAD.
• npm run format — run before committing; whole tree reported unchanged.

UI verification not run: no executable product code, route, component, style, or browser behaviour is touched.

Risk and rollout

• Risk: Low. The change is one new immutable inbox request file. It queues a ledger row and alters no behaviour, no workflow, and no canonical ledger content.
• Rollback: git revert this pull request. Request files are immutable, so the revert simply removes the pending request.
• Provider or production effects: None by this branch. The investigation behind it used read-only GitHub API reads scoped to PR docs(issues): reconcile 38 queued ledger requests into the canonical ledger #2229, explicitly authorised by the user beforehand; no write, re-run, or comment was made, and nothing was pushed to the merged PR.
• RAG impact: none — no file under src/lib/rag/, no retrieval RPC, no ranking configuration, no eval harness and no golden fixture is touched.

Notes

The defect is worth a guard even though it cannot block a merge: the failing check lands on a PR that has already merged, so its only cost is investigative. That cost is real and already measured twice on 2026-08-21 — once in the clarifying comment on #2229, which drew the wrong conclusion about which checks had re-fired, and once in this investigation. The sibling job in the same workflow file, refresh-after-base-push, already scopes itself with pulls.list({ state: "open" }), so the state filter was applied to the push path and omitted from the pull_request_target path rather than considered and rejected.

Repo-wide frequency is unmeasured: GitHub reads were authorised only for PR #2229. It should recur on any pull request whose description a bot edits after merge.

🤖 Generated with Claude Code

https://claude.ai/code/session_01C4SeCeZmgbxbZ4kcKWvtVp

---

Generated by Claude Code

[supabase]

This pull request has been ignored for the connected project sjrfecxgysukkwxsowpy because there are no changes detected in supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your current included review allowance is based on your included PR review attempts over the past 7 days.

Next review available in:46 minutes

Limit details: You’ve used the included review currently available. Your 87 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

How can I continue?

Wait for the limit to reset, then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 88018fa5-0e2f-4ce4-a30f-d65ef6bc062d

📥 Commits

Reviewing files that changed from the base of the PR and between 9b10e59 and 05b63fe.

📒 Files selected for processing (1)

• docs/outstanding-issues-inbox/8dfc4aaf-c22f-45fd-a7f1-0edbd34bf569.json

---

_{Comment @coderabbitai help to get the list of available commands.}