[autohub-codex]

Summary

The full MCP transport comparison is intentionally red for known differences in five tools, so running it on affected PRs adds noise without a usable regression signal. Make it a manual MCP Transport Parity Audit, retaining workflow_dispatch and make test-parity. Manual runs now report failures directly instead of masking them with continue-on-error.

The required CI/node-test suite and deterministic recall contract tests remain unchanged. Document which checks protect PRs and how to run the broader audit. No public API or runtime configuration changes; no breaking changes. Related context: #246.

Review Contract

• The full live parity workflow runs only by manual dispatch, with no PR, push or scheduled trigger.
• Normal PR CI still runs the bridge suite and deterministic recall contract guard.
• Keep the full audit and local make target available; failures in a manually requested audit remain visible.

Validation

• make test: 664 Python tests passed (1 skipped); 43 Node tests passed (5 gated live-audit tests skipped).
• actionlint .github/workflows/mcp-parity.yml: passed.
• Pre-commit checks on both changed files and git diff --check: passed.
• Parsed workflow configuration confirms manual-only audit, no failure masking, and retained PR node-test execution.
• Local Codex review (gpt-6.1-sol, medium): no findings.

Prepared by Codex desktop via AutoJack GitHub App. Babysit to ready only; do not merge.

[jack-arturo]

@codex review

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-22T17:58:27.367454Z	041933f	Manual request

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Delightful!

Reviewed commit: 041933f606

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".

[autohub-codex]

Babysit ready — verified green at 041933f6066ec6273862855847f797ddebd47d62.

• Full live parity audit is manual-only; required deterministic recall contract tests remain in PR CI.
• All six required checks passed: pre-commit, node-test, Python 3.10/3.11/3.12, semantic-pr-title. Optional coverage passed too.
• Local Codex review (gpt-6.1-sol, medium) and hosted Codex review found no issues. Hosted review settled this exact head via reviewed-commit marker; no unresolved threads.
• Size: 63 changed lines across two files. One initial commit; zero remediation pushes or residual findings. One guarded baseline request after automatic-review grace; no repeated requests.
• PR is open, non-draft and mergeable against develop.

Codex desktop via AutoJack GitHub App. No merge performed or auto-merge enabled.