codex-plan-review

Facilitate adversarial AI agent debates to review Markdown implementation plans.

69|23|Updated Feb 20, 2026

Search Tags:#planning #code review #codex #implementation plan #risk assessment #ai debate

Ready to run

npx skills add https://github.com/lploc94/codex_skill --skill codex-plan-review

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: codex-plan-review
Source: https://github.com/lploc94/codex_skill/tree/main/skill-packs/codex-review/skills/codex-plan-review
Command: npx skills add https://github.com/lploc94/codex_skill --skill codex-plan-review

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps ensure your implementation plans are robust, complete, and free of critical flaws before you start coding, preventing costly rework later.

Core Features & Use Cases

Adversarial Review: Simulates a debate between AI agents to rigorously test your plan's logic and identify potential issues.
Plan Quality Improvement: Catches gaps, risks, missing edge cases, and sequencing errors in your development plans.
Use Case: Before starting a new feature, you create a plan.md file. You then run /codex-plan-review to have Claude and Codex debate its contents, ensuring all angles are considered and the plan is solid before implementation begins.

Quick Start

Use the codex-plan-review skill to debate the plan in 'plan.md'.

Frequently Asked Questions about codex-plan-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I review an implementation plan before coding to catch edge cases?▼

To review an implementation plan, you can run an adversarial AI debate between agents to rigorously test your plan's logic, catching gaps, risks, and missing edge cases before coding begins.

What is adversarial AI debate for code review and how does it work?▼

Adversarial AI debate for code review simulates multiple AI agents challenging your implementation plan to identify sequencing flaws and critical issues, ensuring your development plan is robust and complete.

Do I need the Codex CLI to run an AI implementation plan review?▼

Yes, you need the Codex CLI to facilitate the adversarial debate, as the review process requires it to analyze your Markdown plan file and adhere to a strict output format for issue reporting.

How do I use AI agents to assess risks in a Markdown plan file?▼

You provide a Markdown plan file and run the review command to initiate a debate between Claude and Codex, which evaluates the contents to identify risks, gaps, and sequencing errors.

What's the best way to prevent rework on flawed software development plans?▼

The best way to prevent rework is to validate your development plans through an adversarial AI review that identifies critical flaws, missing logic, and sequencing errors before implementation starts.

Can I use this plan review approach for any Markdown plan file?▼

Yes, you can use this approach for any Markdown plan file, provided you have the Codex CLI installed to execute the debate and track issue resolution.