codex

softaworks/agent-toolkit

Run Codex CLI for AI-powered code analysis, refactoring, and automated editing with GPT-5.2.

What is codex?

Codex is a command-line interface for advanced code analysis and automated refactoring powered by GPT-5.2. Use it when you need to analyze code, apply edits, or perform complex software engineering tasks with configurable reasoning effort and sandbox isolation.

Execute code analysis and refactoring tasks via `codex exec` with GPT-5.2 by default
Resume previous Codex sessions to continue analysis or make additional changes
Configure reasoning effort (xhigh, high, medium, low) based on task complexity
Choose sandbox modes (read-only, workspace-write, danger-full-access) for appropriate isolation
Switch between models (gpt-5.2-max, gpt-5.2-mini, gpt-5.1-thinking) for different performance/cost tradeoffs
Run tasks from specified directories with the `-C` flag

How to install codex

npx skills add https://github.com/softaworks/agent-toolkit --skill codex

Prerequisites

Codex CLI v0.57.0 or later installed
OpenAI API access and valid credentials configured
For GPT-5.2 models: appropriate API quota and rate limits

Claude Code

Cursor

Windsurf

Cline

How to use codex

1.Determine the reasoning effort needed (xhigh for ultra-complex, high for complex, medium for standard, low for simple tasks)
2.Select the appropriate sandbox mode: read-only for analysis, workspace-write for edits, danger-full-access for network/broad access
3.Run `codex exec --skip-git-repo-check --sandbox <mode> --config model_reasoning_effort="<level>" 2>/dev/null` with your task description
4.Review the output; stderr is suppressed by default (add without `2>/dev/null` to see thinking tokens if needed)
5.To resume a previous session, use `echo "your new prompt" | codex exec --skip-git-repo-check resume --last 2>/dev/null`

Use cases

Good for

Analyze code quality, architecture, or security issues in read-only mode
Apply automated refactoring or bug fixes with workspace-write sandbox
Resume a previous analysis session to continue with new prompts or deeper investigation
Optimize performance or improve code organization for complex codebases
Execute tasks requiring network access or broad system permissions in danger-full-access mode

Who it's for

Software engineers performing code reviews or refactoring
Development teams automating code analysis and quality checks
Developers needing AI-assisted debugging or architecture analysis
Teams working on complex problem-solving with deep reasoning requirements

codex FAQ

What is the default model and how do I change it?

GPT-5.2 is the default flagship model for software engineering. You can override it with `-m, --model <MODEL>` or use the `/model` slash command within a session. Other options include gpt-5.2-max (ultra-complex tasks), gpt-5.2-mini (cost-efficient), and gpt-5.1-thinking (deep reasoning).

What do the reasoning effort levels mean?

xhigh is for ultra-complex tasks requiring deep analysis; high for complex refactoring/security work; medium for standard tasks like bug fixes; low for simple changes like formatting. Choose based on task complexity.

Can I resume a Codex session later?

Yes. Use `echo "prompt" | codex exec --skip-git-repo-check resume --last 2>/dev/null` to continue a previous session. The resumed session automatically inherits the model, reasoning effort, and sandbox mode from the original.

What sandbox mode should I use?

Use read-only for analysis (default), workspace-write to apply local edits, and danger-full-access only when network or broad system access is required. Always confirm high-impact flags with the user first.

Why is `2>/dev/null` appended to commands?

It suppresses thinking tokens (stderr output) by default for cleaner results. Only remove it if you explicitly want to see the model's reasoning process or need to debug.

Full instructions (SKILL.md)

Source of truth, from softaworks/agent-toolkit.

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. Uses GPT-5.2 by default for state-of-the-art software engineering.

Codex Skill Guide

Running a Task

Default to gpt-5.2 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).
Select the sandbox mode required for the task; default to --sandbox read-only unless edits or network access are necessary.
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>
- --full-auto
- -C, --cd <DIR>
- --skip-git-repo-check
Always use --skip-git-repo-check.
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.
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.
Run the command, capture stdout/stderr (filtered as appropriate), and summarize the outcome for the user.
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
Read-only review or analysis	`read-only`	`--sandbox read-only 2>/dev/null`
Apply local edits	`workspace-write`	`--sandbox workspace-write --full-auto 2>/dev/null`
Permit network or broad access	`danger-full-access`	`--sandbox danger-full-access --full-auto 2>/dev/null`
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	Context window	Key features
`gpt-5.2-max`	Max model: Ultra-complex reasoning, deep problem analysis	400K input / 128K output	76.3% SWE-bench, adaptive reasoning, $1.25/$10.00
`gpt-5.2` ⭐	Flagship model: Software engineering, agentic coding workflows	400K input / 128K output	76.3% SWE-bench, adaptive reasoning, $1.25/$10.00
`gpt-5.2-mini`	Cost-efficient coding (4x more usage allowance)	400K input / 128K output	Near SOTA performance, $0.25/$2.00
`gpt-5.1-thinking`	Ultra-complex reasoning, deep problem analysis	400K input / 128K output	Adaptive thinking depth, runs 2x slower on hardest tasks

GPT-5.2 Advantages: 76.3% SWE-bench (vs 72.8% GPT-5), 30% faster on average tasks, better tool handling, reduced hallucinations, improved code quality. Knowledge cutoff: September 30, 2024.

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: 90% off ($0.125/M tokens) for repeated context, cache lasts up to 24 hours.

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 (--full-auto, --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 Codex CLI v0.57.0 or later for GPT-5.2 model support. The CLI defaults to gpt-5.2 on macOS/Linux and gpt-5.2 on Windows. Check version: codex --version

Use /model slash command within a Codex session to switch models, or configure default in ~/.codex/config.toml.