codex-review-loop

Name: codex-review-loop
Availability: InStock
Author: Soju06

Review pull requests and uncommitted changes with AI, then fix and verify iteratively.

2.6k|399|Updated Jan 7, 2026

Search Tags:#automation #testing #quality assurance #code review #ai #codex cli

Ready to run

npx skills add https://github.com/Soju06/codex-lb --skill codex-review-loop-soju06

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: codex-review-loop
Source: https://github.com/Soju06/codex-lb/tree/main/.agents/skills/codex-review-loop
Command: npx skills add https://github.com/Soju06/codex-lb --skill codex-review-loop-soju06

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the process of code review by leveraging an AI model to identify issues, and then iteratively fixing them until the code meets quality standards.

Core Features & Use Cases

Adversarial Code Review: Uses AI to find bugs, security flaws, and convention violations.
Structured Findings: Parses AI output into actionable, categorized findings.
Automated Fix Loop: Iteratively applies fixes and re-verifies code quality.
Use Case: Integrate this into your CI/CD pipeline to automatically catch and fix common code quality issues before they reach production, ensuring consistent code standards across the team.

Quick Start

Run the codex-review-loop skill to review the current pull request.

Frequently Asked Questions about codex-review-loop

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

FAQPage Schema

How do I automate AI code review for pull requests?▼

Adversarial code review uses an AI model to analyze uncommitted changes or pull requests, identifying bugs, security flaws, and convention violations. It automatically applies approved fixes, re-verifying quality iteratively until the codebase reaches a clean state.

Can I integrate automated code fixes into my CI/CD pipeline?▼

Integrating automated code fixes into CI/CD pipelines catches common code quality issues before production. The adversarial review loop analyzes findings, facilitates human-in-the-loop approval for fixes, and iteratively applies them to maintain consistent code standards.

Do I need Codex CLI to run automated code review?▼

Codex CLI is required to execute the adversarial code review and fix loop. The skill depends on this command-line interface to perform AI analysis, apply atomic fixes, and verify code quality against safety limits.

How does the automated fix loop handle human approval?▼

The automated fix loop facilitates human-in-the-loop approval by parsing AI output into actionable, categorized findings. Users review these structured findings before the system enters an atomic fix-verify-commit loop to apply the approved changes.

What are the limitations of AI code review loops?▼

AI code review loops are bounded by safety limits that halt the process if a clean state is not reached. The adversarial review depends on AI model accuracy and requires adherence to existing project conventions for accurate severity assessment.