Skip to content

test(doc-slop): pin judge codex-home staging contract - #661

Merged
shunk031 merged 1 commit into
mainfrom
fix/doc-slop-review-codex-home
Aug 20, 2026
Merged

test(doc-slop): pin judge codex-home staging contract#661
shunk031 merged 1 commit into
mainfrom
fix/doc-slop-review-codex-home

Conversation

@shunk031

Copy link
Copy Markdown
Owner

Summary

This is a regression-pinning PR, not a production bug fix. Current main already delegates judge-home staging to agent_guidance_eval.initialize_codex_home.

The new tests exercise that path through doc_slop_review.review_with_model and pin the contract:

  • CODEX_HOME, when set, is the fixture source; the default remains ~/.codex.
  • A staged judge home contains only config.toml and auth.json when those source files exist.
  • Source file contents and permission bits are preserved, while unrelated files and directories are not copied.

The investigation also clarified an empirical source of confusion: sqlite state and related files seen in leftover staged homes can be created at runtime by the judge Codex process. Their presence does not prove that the runner copied the source home wholesale. In the live environment, auth.json is absent because authentication uses the configured command helper.

Validation

  • uv run --python 3.14.6 --no-project python -m unittest discover -s tests/python -p 'test_*.py': 93 tests passed.
  • prek run agent-guidance-validate --all-files: passed.
  • prek run agent-guidance-eval --all-files: passed with a cache-local temporary CODEX_HOME and AGENT_GUIDANCE_EVAL_SANDBOX=danger-full-access; temporary state was removed afterward.
  • Complete doc_slop_review.py on this body: passed with model consultation and zero findings.

Coordination

PR #655 changes the judge prompt, and PR #658 changes the deterministic textlint area in the same script. Whichever of those three PRs lands last must update-branch from current main, rerun the complete validation, and re-verify the staging contract.

Do not claim this PR fixes the original bug; it pins the already-correct current-main behavior.

Co-authored-by: Codex <noreply@openai.com>
@shunk031
shunk031 force-pushed the fix/doc-slop-review-codex-home branch from 2c544e1 to ec7f5e5 Compare August 20, 2026 02:34
@shunk031
shunk031 merged commit b3982d1 into main Aug 20, 2026
4 checks passed
@shunk031
shunk031 deleted the fix/doc-slop-review-codex-home branch August 20, 2026 02:38
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant