(chore): pin claude-pr-owner to v0.6.0

[abnegate]

Summary

Bumps the Claude PR orchestrator pin to v0.6.0 and wires a `push_token` secret.

Background:

• v0.5.0 (previous commits in this branch) swapped `improvement` from `/code-review:code-review` to `/skills:improve` so every push gets a fresh review.
• But: CI on Claude's own commits was not running, because the consolidator pushes with the built-in `GITHUB_TOKEN` and GitHub deliberately suppresses `workflow_run`/`synchronize` events for commits pushed that way — an anti-recursion safeguard.

v0.6.0 adds a caller-supplied `push_token` secret. When present, the consolidator uses it for the checkout + push, so the push is treated like a normal user push and Tests/CodeQL/Linter/… re-run on Claude's fixes. When absent, behaviour is unchanged and the consolidator emits a warning.

Action required before merging

Add a repo secret `CLAUDE_PUSH_TOKEN` — a fine-grained PAT (or GitHub App token) scoped to this repo with `contents: write`. Without the secret, v0.6.0 still runs, but CI will keep staying dormant on Claude's pushes and the consolidator will log a warning.

Test plan

• With `CLAUDE_PUSH_TOKEN` set: let Claude push a fix and confirm Tests/CodeQL/Linter are triggered on that commit.
• Without the secret: confirm the consolidator logs a warning and pushes still succeed (just without CI re-trigger).

[coderabbitai]

Warning

Rate limit exceeded

@abnegate has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 11 minutes and 50 seconds before requesting another review.

Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 11 minutes and 50 seconds.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 7954a803-06d4-436e-8fcb-d07fe72e6181

📥 Commits

Reviewing files that changed from the base of the PR and between c87a9a4 and 1d1848c.

📒 Files selected for processing (1)

• .github/workflows/claude.yml

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch chore-bump-claude-pr-owner-0.4.2

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[greptile-apps]

Greptile Summary

This PR updates the claude-pr-owner reusable workflow pin from v0.4.1 (6e883b07) to v0.6.0 (7138fd26) and adds an optional push_token secret (CLAUDE_PUSH_TOKEN). The new secret allows the consolidator to push using a PAT instead of GITHUB_TOKEN, so that GitHub's anti-recursion safeguard no longer suppresses workflow_run/synchronize events on Claude's commits.

Confidence Score: 5/5

Safe to merge; the change is additive, backwards-compatible (secret is optional), and the external workflow is pinned to a verified commit hash.

Single-file CI configuration change with no logic added to the repo itself. No P0/P1 findings; all behaviour is gated on an optional secret that the repo owner controls.

No files require special attention

Important Files Changed

Filename	Overview
.github/workflows/claude.yml	Bumps claude-pr-owner pin from v0.4.1 to v0.6.0 and wires a new push_token secret to allow the consolidator to push commits that re-trigger CI workflows

_{Reviews (3): Last reviewed commit: "(chore): bump claude-pr-owner pin to v0...." | Re-trigger Greptile}

[claude]

CI failure analysis

The failed run (24885064636) on the MySQL adapter tests shows 534 identical errors, all PDOException: SQLSTATE[HY000] [2002] Connection refused at src/Database/PDO.php:29. Every test that tried to open a connection died the same way.

This is infrastructure flakiness — the MySQL service container was not reachable when the test step ran (likely didn't become healthy in time, or crashed mid-run). It is not a code issue.

This branch's only change is a one-line bump to .github/workflows/claude.yml (pinning claude-pr-owner to v0.5.0); it touches no PHP, no adapters, no test config. There is nothing in the diff that could cause MySQL connections to fail.

No code fix needed. Re-running the job should clear it. Exiting without commits.

[claude]

CI Healing Report (Schemaless/MongoDB)

Conclusion: Flaky / pre-existing failure unrelated to this PR. No fix committed.

What failed

Adapter Tests (Schemaless/MongoDB) reported 142 failures and 94 errors. The first failure (test #9) is the root cause:

9) Tests\E2E\Adapter\Schemaless\MongoDBTest::testEvents
Undefined property: stdClass::$cursor
/usr/src/code/src/Database/Adapter/Mongo.php:1224

In Mongo.php:1224:

$result = $this->client->find($name, $filters, $options)->cursor->firstBatch;

The Mongo client occasionally returns a response object without a cursor property, producing the undefined-property error.

Why the cascade

testEvents (tests/e2e/Adapter/Scopes/CollectionTests.php:1465) registers a global EVENT_ALL listener on line 1515 and only unregisters it at lines 1590-1591. Because testEvents errored before reaching the cleanup, the listener leaked into every subsequent test. Each subsequent createDocument/createIndex then triggered the listener, where array_shift($events) returned null and the assertion $this->assertEquals($shifted, $event) failed with messages like Failed asserting that 'document_create' matches expected null. — that is the source of all 141 cascaded failures pointing at CollectionTests.php:1517.

Why this is not from this PR

This branch only modifies .github/workflows/claude.yml (bump claude-pr-owner pin to v0.6.0 and wire push_token). No PHP/test code changed. The same Adapter Tests (Schemaless/MongoDB) job passed on this exact branch in an earlier run on the same SHA family (run 24885064636), confirming the underlying Mongo.php:1224 issue is intermittent and pre-existing.

Recommended follow-up (out of scope for this PR)

• Harden Mongo.php:1224 to handle missing cursor (e.g., null check on the response before accessing ->cursor->firstBatch).
• Wrap the body of testEvents in try/finally so the listener cleanup at lines 1590-1591 always runs even on failure — this would prevent the cascade and make the actual failing test stand out clearly.

Exiting without committing per healing protocol for flaky failures.