skills.rest
mattbutlerengineering
mattbutlerengineeringCOMMUNITY
Explore all skills by this author →
GitHub

codex

Configure OpenAI Codex CLI via repository-scoped .codex/config.toml files.

1|Updated Jan 24, 2026
Search Tags:#cli#codex#configuration#toml#sandbox#mcp servers#approval policy
One-click install
npx skills add https://github.com/mattbutlerengineering/mattbutlerengineering --skill codex-mattbutlerengineering
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/mattbutlerengineering/mattbutlerengineering/tree/main/.claude/skills/codex
Command: npx skills add https://github.com/mattbutlerengineering/mattbutlerengineering --skill codex-mattbutlerengineering

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you set up and consistently use the OpenAI Codex CLI with project-specific configuration so Codex follows the right model, approval workflow, and context files.

Core Features & Use Cases

  • Project configuration via .codex/config.toml: Defines the model, approval policy, sandbox behavior, and which MCP servers Codex can access.
  • Repeatable runs with a shared context: Points Codex to a model instructions file (e.g., AGENTS.md) so answers and edits match your repo standards.
  • Controlled change workflow: Uses approval_policy settings to manage how proposed code changes are handled during execution.
  • Use case: When working in a monorepo, you can run Codex to perform targeted tasks (like reviewing or editing specific parts) while keeping guardrails aligned to your team’s approval and sandbox settings.

Quick Start

Verify Codex CLI is installed by running the Codex executable check, then inspect the existing project config at .codex/config.toml to confirm model and approval policy before running a test command against your repo.

Frequently Asked Questions about codex

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

FAQPage Schema
How do I configure OpenAI Codex CLI for a specific repository?▼

You can configure OpenAI Codex CLI for a specific repository by creating a scoped .codex/config.toml file that defines the target model, approval policy, sandbox behavior, and MCP server access. This ensures Codex runs with your intended project settings.

What is the purpose of an approval policy in Codex config.toml?▼

The approval policy in a Codex config.toml file manages how proposed code changes are handled during execution. Setting an approval policy establishes a controlled change workflow, ensuring proposed edits align with your team's guardrails before being applied.

Can I connect Codex CLI to MCP servers using a TOML configuration?▼

Yes, you can connect Codex CLI to MCP servers by adding optional [mcp_servers] entries in your .codex/config.toml file. This configuration grants Codex access to specified external tools and connectivity during your coding tasks.

How do I set up Codex to follow my repo standards consistently?▼

You can set up Codex to follow repo standards by specifying an optional model_instructions_file, such as AGENTS.md, in your .codex/config.toml. This points Codex to your shared context, ensuring answers and edits match your repository's standards.

Does Codex CLI require sandbox settings to run safely?▼

Sandbox settings are optional in your .codex/config.toml file, but configuring them helps control Codex's execution environment. Defining sandbox behavior ensures the CLI operates within safe boundaries when performing repo-aware coding tasks.

Related Skills

Explore More Software Engineering Skills→
bbsbotbbsbot
1

ln-004-agent-sync

Synchronize AI agent skills and MCP settings across Claude, Gemini CLI, and Codex CLI.

Community
Advanced
alexei-ledalexei-led
35

asking-codex

Delegate coding tasks to an OpenAI Codex agent via the Codex CLI.

Community
Basic
frumu-aifrumu-ai
114

cowork-mcp-config-assistant

Configure MCP servers and replace skill template placeholders with tool names.

Official
Basic
RDCrystalLabRDCrystalLab

tmux-fork

Launch parallel terminal sessions by creating tmux windows for commands and AI coding agents.

Official
Intermediate