codex-skill

Automate codebase modifications with Codex CLI in non-interactive sandbox modes.

322|45|Updated Dec 1, 2025
One-click install
npx skills add https://github.com/Microck/ordinary-claude-skills --skill codex-skill-microck
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-skill
Source: https://github.com/Microck/ordinary-claude-skills/tree/main/codex-skill
Command: npx skills add https://github.com/Microck/ordinary-claude-skills --skill codex-skill-microck

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @openai/codex.

What problem does it solve?

Executing complex coding tasks, from bug fixes to feature implementations, often requires constant human oversight and approval for each step. This Skill leverages the Codex CLI for non-interactive, autonomous execution of AI-powered code analysis and generation tasks.

Core Features & Use Cases

  • Autonomous Code Execution: Implement features, fix bugs, and refactor code from start to finish without requiring approval for every action.
  • Flexible Sandbox Modes: Control execution permissions with read-only, workspace-write, or full-access modes for safe and powerful operations.
  • Model Selection & Configuration: Specify which GPT-5 model to use and configure execution with profiles and custom working directories.
  • Use Case: When you have a well-defined plan to implement a new user authentication feature, use this skill with the --full-auto flag to let Codex autonomously write the code, run tests, and commit the changes, providing a clear summary upon completion.

Quick Start

Use codex-skill with GPT-5.1 in full-auto mode to implement dark mode support for the UI.

Frequently Asked Questions about codex-skill

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

FAQPage Schema
How do I automate code generation and modifications without manual approval at each step?

Autonomous code execution automates feature implementations, bug fixes, and refactoring from start to finish using Codex CLI with GPT-5. Run codex-skill in full-auto mode to let it write code, run tests, and commit changes hands-off, providing a summary when complete.

Can I control execution permissions when using AI-powered code generation?

Yes, codex-skill supports flexible sandbox modes: read-only for analysis without changes, workspace-write for code modifications, or full-access for complete autonomy. Select the mode matching your safety and permission requirements.

What do I need to set up before running automated code tasks with Codex?

Install and configure Codex CLI, select your GPT-5 model version, and optionally specify a working directory or execution profile. The Skill handles the rest non-interactively once prerequisites are met.

How does AI-powered code analysis and generation differ from manual coding workflows?

Code analysis and generation using AI eliminates manual step-by-step approval, allowing complex tasks like multi-file refactoring and feature implementation to complete autonomously while you monitor progress and review the final summary.

Does automated code execution work for testing and dependency updates?

Yes, codex-skill handles feature implementation, bug fixes, refactoring, dependency setup, and automated test generation and execution within your workspace, all without interactive prompts.

What output does the Skill provide after completing an automated code task?

Codex-skill provides structured JSON output, progress streaming during execution, and a final summary report documenting all changes, test results, and commits made during the autonomous task run.