openai/codex-plugin-cc

This system is a development tooling suite centered around a Codex plugin that facilitates the transfer and resumption of Claude sessions within a Codex environment. It includes robust cross-platform support for process management and session import, alongside automated scripts for synchronized version bumping across multiple project manifests. The codebase is supported by an expanded test suite that validates release processes, session handling, and review robustness across different operating systems.

The Codex companion script now supports transferring a Claude session into a Codex thread via the new \transfer\ command, allowing users to resume work in Codex using the provided resume command. Additionally, job listing and task-resumption logic now filters jobs to the current Claude session (identified by the session ID environment variable), ensuring that \resume-last\ and status checks only affect the active session rather than all sessions. The setup and readiness checks have also been updated to rely on the app-server's authentication status rather than the local CLI login status, and the stop-review-gate hook now includes basic error handling to prevent uncaught exceptions.

plugins/codex/scripts · high confidence

The Codex plugin now reliably spawns the app-server on Windows by enabling shell mode and inheriting the system environment, ensuring .cmd shims and Git Bash are resolved correctly. It also introduces a new command to import existing Claude sessions into Codex, allowing users to transfer their conversation history from Claude projects. Additionally, the plugin now gracefully handles older Codex CLI versions that do not support thread naming, and prevents crashes when reviewing untracked directories by skipping them instead of failing.

plugins/codex/scripts/lib · high confidence

A new Node.js script, scripts/bump-version.mjs, has been added to automate version updates across multiple project manifests. This tool ensures consistency by simultaneously updating the version string in package.json, package-lock.json, the Claude Code extension plugin.json, and the marketplace metadata file. It supports bumping to a new version or checking that all manifests are currently aligned, reducing manual effort and preventing version drift during releases.

scripts · high confidence

Added and updated tests in the \tests/\ directory to verify the \bump-version\ script updates all release manifests, the new \codex:transfer\ command delegates sessions to native import, and the \codex:rescue\ command avoids recursion by routing through the Agent tool. The suite also validates that the setup command correctly handles various authentication states (API keys, custom providers, and app-server readiness), that review context collection safely handles special characters in branch names, broken symlinks, untracked directories, and oversized diffs, and that the test infrastructure itself is more portable across platforms and less flaky.

tests · high confidence

The @openai/codex-plugin-cc package has been updated to version 1.0.6. Additionally, new npm scripts 'bump-version' and 'check-version' have been added to support automated version management via a new script file.

(dependencies) · high confidence

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

• CAI 66 → 62 (-3.5)
• Rubric changed (rubric-2026.09.11 → rubric-2026.09.18) — scores are not directly comparable.

• Code Health 74 → 74 (+0.0)
• Architecture 100 → 100 (+0.0)
• Maturity 55 → 55 (+0.0)
• Readiness 63 → 63 (+0.0)
• Security 100 → 100 (+0.0)
• Performance 60 (new)

• Dependency hygiene PARTLY measured — npm pinning read, dependency currency not (no pnpm-resolved versions to grade)
• Documentation: no installation or build instructions (README.md)
• Documentation: no usage examples (README.md)

• Dependency hygiene PARTLY measured — npm pinning read, dependency currency not (the committed lockfile resolved no direct production dependency)

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

openai/codex-plugin-cc was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.