What is the Codex skill for Claude Code?

It is a specialized framework that allows Claude to interface with the Codex CLI for advanced code analysis, automated editing, and refactoring using high-efficiency agentic models.

Which models are supported by the Codex skill?

The skill supports gpt-5.2-codex (default), gpt-5.1-codex-mini for lightweight tasks, and gpt-5.1-codex-max for large-scale, complex operations.

Is it safe to let the skill edit my code?

Yes, the skill utilizes specific sandbox modes like 'read-only' for analysis or 'workspace-write' for edits, and it requires user permission before executing high-impact commands.

Does this skill support Korean language keywords?

Yes, the skill is configured to respond to both English and Korean keywords such as '코덱스', '코드 검증', and 'AI 리뷰' for bilingual developer workflows.

How do I resume a previous Codex session?

You can use the 'codex resume' command or 'codex exec resume --last' to continue a previous session, which automatically inherits the original model and sandbox settings.

Codex CLI Integration

Name: Codex CLI Integration
Author: tygwan

bytygwan

•

Herramientas para Desarrolladores

Automates complex code analysis, refactoring, and intelligent editing using high-performance agentic coding models.

The Codex skill integrates the Ultra Codex framework into Claude Code, enabling developers to perform sophisticated code transformations and deep repository analysis. It provides a structured interface for running the Codex CLI, allowing users to choose between various model tiers like gpt-5.2-codex for general tasks or gpt-5.1-codex-max for large-scale refactoring. With built-in sandbox modes—ranging from read-only reviews to full-access workspace writes—this skill ensures a balance between automation power and security, while also supporting session resumption for efficient, multi-stage development cycles.

Características Principales

01Seamless session resumption with inherited state and context

02Granular control over model reasoning effort for cost and quality optimization

03Configurable sandbox modes including read-only and workspace-write access

04Automated code review and verification powered by agentic AI

05Executes complex code refactoring and analysis via specialized Codex models

061 GitHub stars

Casos de Uso

01Conducting AI-driven code reviews and quality verification checks

02Performing large-scale repository refactoring with high-reasoning agentic models

03Automating repetitive code edits across multiple files in a secure sandbox

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tygwan/ultra-codex-init codex

For use in Claude.ai and ChatGPT

Download Skill