Repository navigation
Conversation
`branches create` intermittently gets a 400 from the platform when the database it configures restarts mid-create (`57P01` / "the database system is shutting down"). The shared live helper now clears any half-made branch and re-issues the create once for that rejection only; a second rejection still fails the test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016f5jekWaBDRtATpEf65mtq
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
Superseded by a newer AI review
🤖 AI Review
Confirmed Claude’s minor error-handling finding: failed cleanup during restart recovery loses the original database-restart error. Codex reported no findings; source inspection found no additional issues. Tests were not run.
Findings
| Severity | Location | Category | Sources | Claim |
|---|---|---|---|---|
| 🟡 MINOR | apps/cli/tests/helpers/branches-live.ts:172 |
error-handling |
claude | If cleanup before retrying branch creation fails, its error replaces the original database-restart error, hiding why recovery began. |
Stats
Claude findings: 1 · Codex findings: 0 · Confirmed: 1 · Refuted: 0 · Uncertain: 0
Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: manual · Workflow run
This review runs once per PR. A maintainer can request another with a /ai-review comment.
There was a problem hiding this comment.
🤖 AI Review
Both independent reviews completed. Code inspection confirms Claude’s two minor findings: retry cleanup can discard the original restart error, and a single LIST check can miss a partially created branch. Codex reported no findings. No additional issues were found; tests were not run.
Findings
Stats
Claude findings: 2 · Codex findings: 0 · Confirmed: 2 · Refuted: 0 · Uncertain: 0
Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: auto · Workflow run
This review runs once per PR. A maintainer can request another with a /ai-review comment.
This branch has not been deployed
2 participants
TL;DR
re-issues the live
branches createonce when the platform rejects it because the database restarted mid-createbranches list(andbranches createonmain/next) failed withunexpected create branch status 400: … FATAL: 57P01: terminating connection due to administrator command(also seen asthe database system is shutting down). the platform restarts the database it is configuring while the create is in flight. the shared live helper now recognises only that rejection, removes the half-made branch if it got listed, and re-issues the create once, the same way test(repo): retry stalled live updates and name stalled analytics polls #7041 handles the network restrictions stall. a second rejection still fails the test, andcreate.live.test.tsuses the same helperref
Generated by Claude Code