Codex

Codex Skill Guide

Running a Task

1. Default to gpt-5.5 model. Ask the user (via AskUserQuestion) which reasoning effort to use (xhigh,high, medium, or low). User can override model if needed (see Model Options below). If gpt-5.5 is unavailable in your environment, fall back to gpt-5.4.
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="<high|medium|low>"
   • --sandbox <read-only|workspace-write|danger-full-access>
   • -a never before exec for non-interactive runs
   • --output-last-message <FILE> when a clean final response file is needed
   • -C, --cd <DIR>
   • --skip-git-repo-check
4. Always use --skip-git-repo-check.
5. When continuing a previous session, use codex exec --skip-git-repo-check resume --last via stdin. When resuming don't use any configuration flags unless explicitly requested by the user e.g. if he species the model or the reasoning effort when requesting to resume a session. Resume syntax: echo "your prompt here" | codex exec --skip-git-repo-check resume --last 2>/dev/null. All flags have to be inserted between exec and resume.
6. IMPORTANT: By default, append 2>/dev/null to all codex exec commands to suppress thinking tokens (stderr). Only show stderr if the user explicitly requests to see thinking tokens or if debugging is needed.
7. Run the command, capture stdout/stderr (filtered as appropriate), and summarize the outcome for the user.
8. After Codex completes, inform the user: "You can resume this Codex session at any time by saying 'codex resume' or asking me to continue with additional analysis or changes."

Quick Reference

Use case	Sandbox mode	Key flags
Non-blocking automated review or analysis	workspace-write	codex -a never exec --sandbox workspace-write --output-last-message <file>
Read-only review or analysis	read-only	codex -a never exec --sandbox read-only --output-last-message <file>
Apply local edits	workspace-write	codex -a never exec --sandbox workspace-write
Permit network or broad access	danger-full-access	codex -a never exec --sandbox danger-full-access
Resume recent session	Inherited from original	echo "prompt" | codex exec --skip-git-repo-check resume --last 2>/dev/null (no flags allowed)
Run from another directory	Match task needs	-C <DIR> plus other flags 2>/dev/null

Model Options

Model	Best for	Notes
gpt-5.5 ⭐	Newest frontier (default): Complex coding, computer use, knowledge work, research workflows	Latest flagship. Available via ChatGPT sign-in and API auth (since 2026-04). Recommended starting point.
gpt-5.4	Flagship alternative for professional work	Available across all Codex platforms. Use as fallback when gpt-5.5 is not in the picker.
gpt-5.4-mini	Fast, efficient mini model for responsive coding tasks and subagents	Lower-cost option for lighter workloads, sub-agents, cleanup tasks.
gpt-5.3-codex	Industry-leading coding model for complex software engineering	Powers GPT-5.4's coding abilities. Use when coding-specific quality is paramount.
gpt-5.3-codex-spark	Real-time coding iteration (text-only research preview)	ChatGPT Pro subscribers only. Optimized for near-instant coding iteration.

Selection guide: Start with gpt-5.5 for most tasks. Drop to gpt-5.4 when 5.5 isn't available. Use gpt-5.3-codex when you need maximum coding quality without 5.5 reasoning overhead. Use gpt-5.4-mini for sub-agents and cleanup workloads.

Verify exact context window / pricing on the official model card — values may shift as preview/stable status changes.

Reasoning Effort Levels:

• xhigh - Ultra-complex tasks (deep problem analysis, complex reasoning, deep understanding of the problem)
• high - Complex tasks (refactoring, architecture, security analysis, performance optimization)
• medium - Standard tasks (refactoring, code organization, feature additions, bug fixes)
• low - Simple tasks (quick fixes, simple changes, code formatting, documentation)

Cached Input Discount: Repeated context is heavily discounted with cache lasting up to 24 hours. Verify current rates on the official OpenAI pricing page.

Following Up

• After every codex command, immediately use AskUserQuestion to confirm next steps, collect clarifications, or decide whether to resume with codex exec resume --last.
• When resuming, pipe the new prompt via stdin: echo "new prompt" | codex exec resume --last 2>/dev/null. The resumed session automatically uses the same model, reasoning effort, and sandbox mode from the original session.
• Restate the chosen model, reasoning effort, and sandbox mode when proposing follow-up actions.

Error Handling

• Stop and report failures whenever codex --version or a codex exec command exits non-zero; request direction before retrying.
• Before you use high-impact flags (--sandbox danger-full-access, --skip-git-repo-check) ask the user for permission using AskUserQuestion unless it was already given.
• When output includes warnings or partial results, summarize them and ask how to adjust using AskUserQuestion.

CLI Version

Requires a recent Codex CLI version for GPT-5.5 model support. Check version: codex --version and update with npm install -g @openai/codex@latest if needed. Use /model slash command within a session to switch, or set default in ~/.codex/config.toml.