What are the different sandbox modes available?

The skill supports 'read-only' for analysis, 'workspace-write' for code modifications, and 'danger-full-access' for system-level operations.

What is the primary benefit of the Codex skill for Claude Code?

It enables Claude Code to utilize the OpenAI Codex CLI for deeper code reasoning, complex refactoring, and specialized execution modes that are not available in standard environments.

Can I adjust how much effort the model puts into reasoning?

Yes, you can specify reasoning levels from 'none' for low-latency tasks to 'xhigh' for extremely complex architectural or algorithmic problems.

How does this skill handle restricted HPC or Slurm environments?

The skill detects HPC environments and uses the --yolo flag to bypass Landlock sandbox restrictions, allowing the CLI to function correctly on restricted cluster nodes.

Is it possible to continue a previous coding session?

Yes, the skill includes a 'resume' capability that automatically inherits the model selection, reasoning effort, and configuration from your last active session.

OpenAI Codex CLI Integration

Name: OpenAI Codex CLI Integration
Author: sypsyp97

bysypsyp97

•

Developer Tools

Integrates the OpenAI Codex CLI into Claude Code for deep code analysis, automated refactoring, and specialized HPC environment execution.

About

The Codex skill empowers Claude Code with the advanced capabilities of the OpenAI Codex CLI, enabling sophisticated code transformations and architectural analysis. It is specifically optimized for professional developer workflows, offering granular control over model reasoning effort levels and sandbox permissions. A standout feature is its native support for High-Performance Computing (HPC) and Slurm environments, where it intelligently bypasses Landlock sandbox restrictions to ensure reliable execution. Whether you are performing security audits, complex refactoring, or automated documentation, this skill provides the necessary tools to handle high-stakes code modifications with precision.

Key Features

Persistent session management with the ability to resume and extend previous tasks
Configurable execution sandboxes for read-only, workspace-write, or full-access tasks
Granular reasoning effort controls ranging from 'low' to 'xhigh' depth
Advanced code analysis and automated refactoring via OpenAI Codex CLI
Specialized HPC/Slurm support with Landlock sandbox bypass (YOLO mode)
26 GitHub stars

Use Cases

Refactoring large-scale legacy codebases to modern patterns on remote HPC clusters
Conducting automated security audits and architectural reviews with high-reasoning models
Executing non-interactive, automated code edits within protected development environments

About

Key Features

Persistent session management with the ability to resume and extend previous tasks
Configurable execution sandboxes for read-only, workspace-write, or full-access tasks
Granular reasoning effort controls ranging from 'low' to 'xhigh' depth
Advanced code analysis and automated refactoring via OpenAI Codex CLI
Specialized HPC/Slurm support with Landlock sandbox bypass (YOLO mode)
26 GitHub stars

Use Cases

Refactoring large-scale legacy codebases to modern patterns on remote HPC clusters
Conducting automated security audits and architectural reviews with high-reasoning models
Executing non-interactive, automated code edits within protected development environments