fix(onboard): preserve rerun config migrations#87328
Conversation
|
Codex review: needs real behavior proof before merge. Reviewed May 27, 2026, 1:01 PM ET / 17:01 UTC. Summary PR surface: Source +134, Tests +359, Other +162. Total +655 across 19 files. Reproducibility: no. high-confidence live reproduction was run in this read-only review. The migration edge is source-reproducible from current main because ordinary non-interactive onboard writes bypass the pending install-record helper while this PR routes those writes through it. Review metrics: 2 noteworthy metrics.
Merge readiness Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch. Rank-up moves:
Proof guidance: Risk before merge
Maintainer options:
Next step before merge Security Review detailsBest possible solution: Land the migration isolation only after maintainer review confirms the compatibility boundary and the PR body includes redacted real onboard rerun proof for the updated head. Do we have a high-confidence way to reproduce the issue? No high-confidence live reproduction was run in this read-only review. The migration edge is source-reproducible from current main because ordinary non-interactive onboard writes bypass the pending install-record helper while this PR routes those writes through it. Is this the best way to solve the issue? Unclear as merge-ready, but the code direction is plausible: migrating legacy install records before the final protected config write is the narrow maintainable shape. The missing piece is updated-head real behavior proof and maintainer acceptance of the compatibility and automation risks. AGENTS.md: found and applied where relevant. Codex review notes: model gpt-5.5, reasoning high; reviewed against 11dfef201f81. Label changesLabel changes:
Label justifications:
Evidence reviewedPR surface: Source +134, Tests +359, Other +162. Total +655 across 19 files. View PR surface stats
What I checked:
Likely related people:
What the crustacean ranks mean
Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics. How this review workflow works
|
|
ClawSweeper PR egg 🎁 Pass real behavior proof to wake the egg and unlock a hatchable treat. Where did the egg go?
|
|
No dependency changes detected. Learn more about Socket for GitHub. 👍 No dependency changes detected in pull request |
This reverts commit d66c290.
d66c290 to
24c20d3
Compare
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 24c20d334d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| } | ||
| const toolingTargets = groupedTargets.get("tooling") ?? []; | ||
| if ( | ||
| !watchMode && |
There was a problem hiding this comment.
Run the isolated tooling test in watch plans
When pnpm test --watch test/scripts (or a similar broad test/scripts watch target) is used, this guard prevents adding the new tooling-isolated plan, while test/vitest/vitest.tooling.config.ts now excludes test/scripts/openclaw-e2e-instance.test.ts. I checked the routing in scripts/test-projects.test-support.mjs; the result is that the helper test no longer runs at all in that watch workflow, whereas it was included before this split.
Useful? React with 👍 / 👎.
|
Verification for head Behavior addressed: non-interactive and wizard onboarding reruns preserve existing
|
Summary
Verification
Refs #84748
Closes #84692
Co-authored-by: yetval [email protected]