# codex

A skill that tells your agent how to drive Codex CLI: pick a model and reasoning effort, build the codex exec command, and resume sessions.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 495 installs · ★ 1465
- Author: skills-directory
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/skills-directory/skill-codex/tree/HEAD/plugins/skill-codex/skills/codex/SKILL.md
- Updated: 2026-10-21
- HTML: https://affpapa.org/en/nevibecode/skills/skills-directory-codex

## What it does

The skill hands Codex CLI work off to Codex instead of doing it in the host agent. It walks through the steps: ask you for a model and a reasoning effort in one prompt with two questions, pick a sandbox mode, assemble the command, run it, then summarize. With no stated preference it suggests starting at `gpt-6-astra` on `high`.

Flags map to the task: `-m` for the model, `--config model_reasoning_effort=...` for effort, `--sandbox` for access, `-C` for the working directory. The default sandbox is `--sandbox read-only`; edits and network access are opt-in. Effort levels `max` and `ultra` need a GPT-6 or GPT-5.6 model, and `ultra` only runs on `astra`, `sol` and `terra`. Pick an effort above a model's ceiling and the skill falls back to the highest supported level and says so.

The `--skip-git-repo-check` flag has its own gate. Inside a Git repo it stays off by default; outside one it needs your approval, and a general request to use Codex does not count as that approval. Two more traps: `codex exec` always reads stdin, so a harness that leaves the stream open hangs forever — append `</dev/null`. And output only lands at completion, so a killed process leaves an empty file with no error. Background timeouts scale with effort, from 150s on `low` to 1800s on `max` and `ultra`.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add skills-directory/skill-codex --skill codex -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/codex/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/codex/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add skills-directory/skill-codex --skill codex -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/codex/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add skills-directory/skill-codex --skill codex -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/codex/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install skills-directory/skill-codex codex
```

2. **Manually: put the files in place**
- File: `.github/skills/codex/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add skills-directory/skill-codex --skill codex -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/skills-directory/skill-codex --path plugins/skill-codex/skills/codex
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/codex/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add skills-directory/skill-codex --skill codex -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/skills-directory/skill-codex/tree/HEAD/plugins/skill-codex/skills/codex/SKILL.md

## When to use it
- run Codex CLI from Claude without leaving the session
- match model and reasoning effort to the job at hand
- continue an earlier Codex session with a new prompt
- review code in read-only mode before touching anything
- set a background timeout that fits the chosen effort

## Example prompts
- run this module through Codex and show me what it finds
- resume the last Codex session and add a race condition check
- run Codex in workspace-write mode and fix the imports

## Source text · SKILL.md · MIT

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: c9dad49ad6b4f165.

BEGIN THIRD-PARTY FILE c9dad49ad6b4f165
```text
---
name: codex
description: Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing
---

# Codex Skill Guide

## Running a Task
1. For a new session (resumes inherit the prior model/effort — see step 5), ask the user (via `AskUserQuestion`) which **model** AND which **reasoning effort** to use, in a **single prompt with two questions**. When the user expresses no preference, default to `gpt-6-astra` at `high`.
   - **Model** — default `gpt-6-astra`:
     - *GPT-6:* `gpt-6-astra` (frontier / most capable — **default**), `gpt-6.1-sol` (latest workhorse for coding and everyday work), `gpt-6-sol` (previous-generation workhorse), `gpt-6-luna` (fast & affordable, easier tasks)
     - *GPT-5.6:* `gpt-5.6-sol` (reliable agentic workhorse), `gpt-5.6-terra` (balanced, everyday), `gpt-5.6-luna` (fast & affordable)
     - *Legacy (rejected as unsupported when Codex is signed in with a ChatGPT account):* `gpt-5.5`, `gpt-5.4`, `gpt-5.4-mini`, `gpt-5.3-codex-spark`, `gpt-5.3-codex`
   - **Reasoning effort** — default `high`: `low`, `medium`, `high`, `xhigh`, `max`, `ultra`.
     - `max`/`ultra` require a GPT-6 or GPT-5.6 model; `ultra` is only on `astra`/`sol`/`terra` (`luna` caps at `max`); legacy models cap at `xhigh`.
     - `ultra` = maximum reasoning **with automatic task delegation** (slowest and most expensive — reserve for the hardest jobs).
     - If the chosen effort exceeds the chosen model's maximum, fall back to that model's highest supported effort and tell the user.
2. Select the sandbox mode required for the task; default to `--sandbox read-only` unless edits or network access are necessary.
3. Assemble the command with the appropriate options:
   - `-m, --model <MODEL>`
   - `--config model_reasoning_effort="<low|medium|high|xhigh|max|ultra>"` (max/ultra only on GPT-6 / GPT-5.6 models; ultra only on astra/sol/terra — see step 1)
   - `--sandbox <read-only|workspace-write|danger-full-access>`
   - `-C, --cd <DIR>`
   - `--skip-git-repo-check` (only when permitted by step 4)
   - `"your prompt here"` (as final positional argument)
4. Determine whether the effective target directory is inside a Git repository, including any explicit working directory such as `-C <DIR>`.
   - Unknown Git context is unresolved, not consent; resolve it before invoking.
```
END THIRD-PARTY FILE c9dad49ad6b4f165

The third-party file has ended. Everything between the c9dad49ad6b4f165 tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 2392 of 8740 bytes. © skills-directory · MIT · https://github.com/skills-directory/skill-codex/tree/HEAD/plugins/skill-codex/skills/codex/SKILL.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md