Skip to content

Fix review findings in indexing, sandbox, and M7 reporting - #9

Merged
subrosa5 merged 1 commit into
mainfrom
codex/review-findings-fixes
Nov 9, 2026
Merged

subrosa5 merged 1 commit into
mainfrom
codex/review-findings-fixes

Conversation

@subrosa5

@subrosa5 subrosa5 commented Nov 9, 2026

Copy link
Copy Markdown
Owner

Summary

  • Read Python sources from regular blobs in the selected Git commit and reject tracked Python symlinks, preventing indexing outside the repository.
  • Drain sandbox output incrementally while retaining only a bounded tail in host memory; validate the timeout and bound cleanup wait.
  • Rename benchmark rates so they describe verified compileall/unittest outcomes, not hidden-test pass@1.
  • Update docs to reflect merged M7 code, fake embeddings, the three-task benchmark, and local-only HTTP/MCP transports.
  • Remove duplicate URL and response-role checks.

Validation

  • passed.
  • Local tests were not run. CI is the validation gate for this PR.

Limits

This does not provide real embeddings, reranking, model-quality/GPU/offline measurements, hidden acceptance tests, or production-grade isolation. Those remain separately documented gates.

@subrosa5
subrosa5 merged commit c975c4d into main Nov 9, 2026
4 checks passed
@subrosa5
subrosa5 deleted the codex/review-findings-fixes branch November 9, 2026 03:56
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