Does the Codex skill support workspace edits?

Yes, it features a 'workspace-write' sandbox mode specifically for generating code and applying edits directly to your project files.

Which AI models can I use with the Codex skill?

It supports several models via JetBrains AI staging, including GPT-4o-2024-11-20 for speed and O1-2024-12-17 or O3-mini-2025-01-31 for deep reasoning tasks.

What is the Codex skill for Claude Code?

The Codex skill is an integration that allows Claude to delegate complex or repetitive coding tasks to the JetBrains Codex CLI, utilizing specialized OpenAI models optimized for code.

How do I resume a previous Codex session?

You can use the 'resume' command within the codex exec utility to pick up a previous task exactly where it left off, maintaining context for ongoing refactors.

When should I use Codex instead of native Claude?

Use Codex for mechanical tasks involving many files, such as boilerplate generation or large refactors. Keep Claude for strategic planning and interactive problem-solving.

Codex CLI Bridge

Name: Codex CLI Bridge
Author: ashchupliak

byashchupliak

•

Productividad y Flujo de Trabajo

Delegates repetitive coding tasks and large-scale refactoring to the JetBrains Codex CLI for efficient parallel processing.

Acerca de

The Codex skill acts as a high-performance bridge between Claude Code and the JetBrains AI ecosystem, allowing users to offload mechanical, repetitive, or high-volume tasks to the Codex CLI. By leveraging specialized models like GPT-4o and O3-mini, this skill excels at codebase-wide refactoring, boilerplate generation, and deep security analysis across multiple files. It enables a hybrid workflow where Claude handles strategic architectural decisions while Codex manages the heavy lifting of implementation, pattern-based transformations, and exhaustive documentation.

Características Principales

High-reasoning mode for complex logic audits and security reviews
Configurable sandboxing with read-only and workspace-write modes
Multi-model integration with GPT-4o, O1, and O3-mini via JetBrains AI
Parallel processing for large-scale codebase refactoring and analysis
1 GitHub stars
Session persistence allowing for task resumption and continuation

Casos de Uso

Performing repetitive architectural migrations or pattern updates across hundreds of files
Conducting deep-dive security vulnerability assessments using O-series reasoning models
Generating comprehensive unit test suites for entire service modules simultaneously

Acerca de

Características Principales

High-reasoning mode for complex logic audits and security reviews
Configurable sandboxing with read-only and workspace-write modes
Multi-model integration with GPT-4o, O1, and O3-mini via JetBrains AI
Parallel processing for large-scale codebase refactoring and analysis
1 GitHub stars
Session persistence allowing for task resumption and continuation

Casos de Uso

Performing repetitive architectural migrations or pattern updates across hundreds of files
Conducting deep-dive security vulnerability assessments using O-series reasoning models
Generating comprehensive unit test suites for entire service modules simultaneously