codex-review-workflow

Automated code review uses OpenAI Codex CLI to scan your codebase for security vulnerabilities, bugs, and best practice violations in iterative cycles. The Skill runs multiple review passes on your Git repository, flagging issues like SQL injection, weak password handling, and maintainability problems, then tracks fixes across iterations until code passes validation.

Yes. This Skill integrates directly into Git workflows by operating on your repository and tracked files. It requires Codex CLI on PATH and an initialized Git repository, then automates six-step validation cycles including initial review, issue tracking, iterative fixes, and final status reporting within your existing development process.

You need OpenAI Codex CLI installed and available on PATH, an initialized Git repository, and explicit review prompts defining what to validate. The Skill then automates the remaining workflow—no additional setup required beyond these three prerequisites.

Iterative review runs multiple fix-and-review cycles automatically, allowing the system to validate initial feedback, track corrections, and re-check modified code. This loop catches cascading issues and ensures fixes don't introduce new problems, stopping at iteration limits or when code passes all validation gates.

Use automated review for consistent, repeatable validation of security, bugs, and best practices across feature branches and pull requests—particularly when scaling teams or enforcing standards. Manual review remains valuable for architectural decisions and domain-specific logic; this Skill handles the systematic scanning layer.