Facilitates an iterative debate loop between Claude and Codex to reach consensus on code reviews or freeform technical questions.
Install
mkdir -p .claude/skills/codex-debate && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/19082" && unzip -o skill.zip -d .claude/skills/codex-debate && rm skill.zipInstalls to .claude/skills/codex-debate
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Run an automated codex⇄Claude debate to consensus — no round cap, no deadlock exit. Two explicit subcommands. `review` (also the bare/back-compat default) — codex critiques the current diff and Claude (author) fixes/disputes, looping until they agree, and the trail is committed + posted to the PR. `answer` — Claude and codex each answer a freeform prompt, then cross-check until they agree, and a unified answer is returned. The debate runs over a LIVE codex session in a split terminal beside you (driven per the /kolu skill), not a headless workflow. Use when the user types `/codex-debate`, asks to "have codex review this", "run the codex debate", "review this PR with codex", "argue this with codex until you agree", or passes a question to "have Claude and codex debate/answer until they agree".Key capabilities
- →Orchestrate an automated debate between Codex and Claude
- →Review code diffs iteratively until consensus is reached
- →Answer freeform prompts through cross-checking until agreement
- →Commit debate trails and post to pull requests in review mode
- →Return a unified answer and saved transcript in answer mode
- →Operate with no round cap and no deadlock exit
How it works
This skill orchestrates a debate between Codex and Claude, where they iteratively review code or answer prompts, reaching consensus through a live, interactive session in a split terminal.
Inputs & outputs
When to use codex-debate
- →Reviewing a pull request collaboratively with an AI peer
- →Resolving technical design disagreements via consensus
- →Validating code improvements through cross-checking
- →Debugging logic by comparing two AI perspectives
About codex-debate
This tool orchestrates a debate between Claude and Codex within your terminal. It offers two modes: a 'review' mode for iterative code diff consensus and an 'answer' mode for resolving technical questions through collaborative reasoning.
Run an automated codex⇄Claude debate to consensus — no round cap, no deadlock exit. Two explicit subcommands. `review` (also the bare/back-compat default) — codex critiques the current diff and Claude (author) fixes/disputes, looping until they agree, and the trail is committed + posted to the PR. `
When not to use it
- →When a mutating action is triggered by freeform prose
- →When the user expects a headless workflow without a live Codex session
- →When the user cannot spawn a sibling split terminal for Codex
Limitations
- →Requires running inside a kolu terminal with a sibling split
- →The reviewer runs unsandboxed with `--yolo`
- →Terminal references must survive kaval re-keying
How it compares
This workflow automates an iterative debate with a live AI, providing a structured consensus-building process that differs from manual code reviews or single-agent question answering.
Compared to similar skills
codex-debate side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| codex-debate (this skill) | 0 | 10d | Review | Advanced |
| claude-automation-recommender | 47 | 2mo | Review | Beginner |
| subagent-driven-development | 14 | 7mo | No flags | Advanced |
| coding-agent | 16 | 2mo | Review | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by juspay
View all by juspay →You might also like
claude-automation-recommender
anthropics
Analyze a codebase and recommend Claude Code automations (hooks, subagents, skills, plugins, MCP servers). Use when user asks for automation recommendations, wants to optimize their Claude Code setup, mentions improving Claude Code workflows, asks how to first set up Claude Code for a project, or wants to know what Claude Code features they should use.
subagent-driven-development
davila7
Use when executing implementation plans with independent tasks in the current session
coding-agent
openclaw
Run Codex CLI, Claude Code, OpenCode, or Pi Coding Agent via background process for programmatic control.
requesting-code-review
obra
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
omo
cexll
Use this skill when you see `/omo`. Multi-agent orchestration for "code analysis / bug investigation / fix planning / implementation". Choose the minimal agent set and order based on task type + risk; recipes below show common patterns.
parallel-agents
davila7
Multi-agent orchestration patterns. Use when multiple independent tasks can run with different domain expertise or when comprehensive analysis requires multiple perspectives.