Menu

Individual skill

claude-codex-pair

0upvotes

Coordinate Claude Code and Codex on the same software task with one active writer, an independent read-only reviewer, explicit ownership transfers, evidence-based handoffs, and safe use of the official Codex plugin for Claude Code. Use when the user asks Claude and Codex to work together, delegate to each other, review or challenge each other's work, transfer a session, rescue a task, or continue work across both agents. Do not use for ordinary single-agent implementation or factual questions.

skills/claude-codex-pair/SKILL.md

Read pinned source

Source commit b0f612844e1aa7d1f2817454d1a5fb60457b8a52

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
No
Repository scan complete
No
Skill scan complete
Yes
Observed behavior
Not checked yet
Last checked
Nov 6, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
3
Repository last updated
Nov 1, 2026
License
MIT

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# Claude–Codex Pair

Treat Claude Code and Codex as a writer–challenger pair. The roles rotate by task; neither
agent is permanently the lead. Optimize for independent evidence, not agreement.

## Non-Negotiable Invariant

Exactly one agent owns write access to the current worktree at a time.

- The **writer** may inspect, edit, test, and create local commits within the accepted scope.
- The **challenger** is read-only. It attempts to falsify the writer's result with code,
  tests, specs, logs, documentation, and reproducible commands.
- A review verdict never grants permission to edit.
- Do not run two write-capable agents concurrently in the same worktree.