Which models can I use with this skill?

The skill supports gpt-5.1, gpt-5.1-codex, gpt-5.1-codex-max, and gpt-5.1-mini variants.

What is the Codex CLI skill for Claude Code?

It is a specialized skill that enables Claude to interface with the Codex CLI to perform tasks like code analysis, refactoring, and automated editing using advanced AI models.

Is it safe to run automated edits with this skill?

The skill uses a sandbox system (read-only, workspace-write, or danger-full-access) and is programmed to ask for user permission before executing high-impact or full-auto commands.

What is 'Reasoning Effort' in Codex CLI?

It is a setting that determines how much computational 'thought' the model applies to a task, configurable as low, medium, or high to balance speed and accuracy.

Can I continue a previous Codex session?

Yes, the skill provides a specific command to resume the last session, which automatically inherits the model and sandbox settings from the original run.

Codex CLI Integration

Name: Codex CLI Integration
Author: rafaelcalleja

byrafaelcalleja

•

Developer Tools

Executes and manages OpenAI Codex tasks for advanced code analysis, automated refactoring, and intelligent editing within secure sandboxes.

About

The Codex CLI skill allows Claude to leverage powerful OpenAI Codex models to perform complex development tasks directly within your environment. It provides a structured interface for selecting specific GPT-5.1 model variants, adjusting reasoning effort, and managing sandbox security levels—ranging from read-only analysis to full workspace writes. With built-in support for session resumption and autonomous execution modes, this skill bridges the gap between Claude's reasoning and the specialized code-manipulation capabilities of the Codex ecosystem.

Key Features

Granular sandbox security modes: read-only, workspace-write, and full-access
Configurable reasoning effort (low, medium, high) for cost and quality control
1 GitHub stars
Multi-model support including gpt-5.1-codex and gpt-5.1-mini
Automated execution with full-auto mode for rapid refactoring
Session resumption functionality to continue previous coding tasks seamlessly

Use Cases

Performing deep refactoring of legacy code using high-reasoning Codex models
Running automated code analysis and security reviews in a read-only sandbox
Automating repetitive codebase edits across multiple directories with workspace-write permissions

About

Key Features

Granular sandbox security modes: read-only, workspace-write, and full-access
Configurable reasoning effort (low, medium, high) for cost and quality control
1 GitHub stars
Multi-model support including gpt-5.1-codex and gpt-5.1-mini
Automated execution with full-auto mode for rapid refactoring
Session resumption functionality to continue previous coding tasks seamlessly

Use Cases

Performing deep refactoring of legacy code using high-reasoning Codex models
Running automated code analysis and security reviews in a read-only sandbox
Automating repetitive codebase edits across multiple directories with workspace-write permissions