skills.rest
kazuph
kazuphCOMMUNITY
Explore all skills by this author →
GitHub

codex

Execute AI-driven code analysis, implementation, and review tasks via Codex CLI.

Updated Aug 23, 2026
Search Tags:#developer tools#code review#debugging#code generation#codex#ai coding
One-click install
npx skills add https://github.com/kazuph/dotfiles --skill codex-kazuph
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/kazuph/dotfiles/tree/main/plugins/kazuph-dotfiles/skills/codex
Command: npx skills add https://github.com/kazuph/dotfiles --skill codex-kazuph

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill streamlines the software development lifecycle by leveraging an advanced AI model (Codex CLI) to assist with code analysis, feature implementation, and quality assurance through automated reviews.

Core Features & Use Cases

  • Code Investigation: Analyze existing codebases for understanding, debugging, or identifying potential issues without making changes.
  • Code Implementation: Generate new code, refactor existing code, or implement new features based on natural language prompts, with the ability to modify files directly.
  • Code Review: Automatically review uncommitted changes, branch differences, or specific commits against custom criteria or predefined models.
  • Use Case: A developer needs to understand a complex function. They can use /codex investigate "Explain this function and its dependencies" to get a detailed analysis. Later, they might use /codex implement "Add error handling for network requests" to modify the code, and finally /codex review to check their changes.

Quick Start

Use the codex skill to investigate the current codebase with the prompt "Analyze the security vulnerabilities in the authentication module".

Frequently Asked Questions about codex

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

FAQPage Schema
How do I perform an automated code review on uncommitted changes?▼

Automated code review of uncommitted changes is executed by running the Codex CLI against your current branch diff to identify potential issues. It analyzes the modifications and provides feedback based on predefined models or custom criteria.

Can I use natural language prompts to implement new features and modify files directly?▼

Natural language prompts can be used to implement new features or fixes, allowing the AI to modify files directly. The implementation process analyzes your request, generates the necessary code, and applies the changes to your codebase.

What is the best way to investigate a complex codebase for debugging?▼

Investigating a complex codebase for debugging is done by executing an AI-driven analysis without making changes. This process examines the existing code to explain functions, dependencies, and identify potential issues to aid understanding.

Does the AI code generation tool require a specific environment to run?▼

AI code generation requires a TTY environment for execution. Specific Codex CLI flags are necessary for model selection, sandboxing, and output control to ensure the code analysis and implementation tasks run correctly.

Why use AI for code analysis instead of manual review?▼

AI code analysis streamlines the software development lifecycle by automating code investigation, implementation, and quality assurance. It leverages advanced models to analyze codebases and implement fixes faster than manual review.

Can I review specific commits rather than just uncommitted changes?▼

Reviewing specific commits is supported alongside uncommitted changes and branch diffs. The automated code review process can target specific points in your git history to evaluate modifications against your criteria.

Related Skills

Explore More Software Engineering Skills→
AlekseiULAlekseiUL
16

cursor-agent

Run Cursor CLI agent tasks in tmux-backed environments.

Community
Intermediate
benchflow-aibenchflow-ai
317

branch-review

Run linting, type checks, and tests on git branch diffs before pushing.

Official
Advanced
FFCfelps1FFCfelps1

ce-work-beta

Delegate code implementation tasks to external systems like Codex.

Community
Advanced
yash-gargyash-garg

design-an-interface

Generate and compare alternative software interface designs for module decisions.

Community
Intermediate