Skills by agent

Codex skills

867 audited skills whose own text says they work in Codex. Each one carries our verdict, so you can see what a skill can touch before you install it: 474 SAFE, 161 CAUTION, 232 BLOCK. “Works in” is the skill’s claim; we have not tested it.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Codex skills”. Retrieved 28 Oct 2026. https://modelsatlas.com/skills/for/codex

1 · Pick a skill

Prefer SAFE. Open CAUTION skills and read the findings first.

2 · Check what it touches

Each page lists the filesystem, network and shell use the audit observed in the source.

3 · Pin the commit

Our verdict covers the audited commit only. Each page shows it.

Verdicts across these 867 skills

✓SAFE474!CAUTION161✕BLOCK232

How these skills say to install

427 of 867 document any install command. Commands are as the skills’ repositories show them: we list them, we never run them.

867 skills for Codex · page 1 of 37 · ranked by trust score, then stars

Skills For Real Engineers

mattpocock/skills

✓SAFE B

Ask which skill or flow fits your situation. A router over the skills in this repo.

★ 305k89/100

Brainstorming

obra/superpowers

✓SAFE B

You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).

★ 303k89/100

Skills For Real Engineers

mattpocock/skills

✓SAFE B

Ask which skill or flow fits your situation. A router over the skills in this repo.

★ 302k89/100

Ask Matt

mattpocock/skills

✓SAFE B

Ask which skill or flow fits your situation. A router over the skills in this repo.

★ 301k89/100

Everything Claude Code

affaan-m/ECC

✓SAFE B

Development conventions and patterns for everything-claude-code. JavaScript project with conventional commits.

★ 265k89/100

Safety Guard

affaan-m/ECC

✓SAFE B

本番システムでの作業時や、エージェントを自律的に実行する際に破壊的な操作を防ぐためにこのスキルを使用してください。

★ 265k89/100

Video Editing

affaan-m/ECC

✓SAFE B

実写素材のカット、構築、強化のためのAI支援ビデオ編集ワークフロー。生の撮影素材からFFmpeg、Remotion、ElevenLabs、fal.aiを経て、DescriptまたはCapCutで最終仕上げを行う完全なパイプラインをカバーする。ユーザーがビデオの編集、素材のカット、vlogの作成、またはビデオコンテンツの構築を望む場合に使用する。

★ 265k89/100

Workspace Surface Audit

affaan-m/ECC

✓SAFE B

アクティブなリポジトリ、MCPサーバー、プラグイン、コネクター、環境サーフェス、ツールのセットアップを監査し、最も価値の高いECCネイティブスキル、フック、エージェント、オペレーターワークフローを推奨する。ユーザーがClaude Codeのセットアップを支援してほしい場合や、環境で実際に何が使えるかを理解したい場合に使用する。

★ 265k89/100

Documentation Lookup

affaan-m/ECC

✓SAFE B

Use up-to-date library and framework docs via Context7 MCP instead of training data. Activates for setup questions, API references, code examples, or when the user names a framework (e.g. React, Next.js, Prisma).

★ 263k89/100

Ecc Guide

affaan-m/ECC

✓SAFE B

ECC の現在のエージェント、スキル、コマンド、フック、ルール、インストールプロファイル、およびプロジェクトオンボーディングをガイドしています。ライブリポジトリサーフェスを読んでから回答するようユーザーをガイドします。

★ 263k89/100

Knowledge Ops

affaan-m/ECC

✓SAFE B

複数のストレージレイヤー(ローカルファイル、MCP メモリ、ベクターストア、Git リポジトリ)にわたるナレッジベースの管理、取り込み、同期、検索。ユーザーが知識システム全体で保存・整理・同期・重複排除・検索を行いたい場合に使用します。

★ 263k89/100

Agent Eval

affaan-m/ECC

✓SAFE B

カスタムタスクでコーディングエージェント(Claude Code、Aider、Codex など)をヘッドツーヘッドで比較し、合格率、コスト、時間、一貫性のメトリクスを測定します

★ 262k89/100

Ai Regression Testing

affaan-m/ECC

✓SAFE B

AI 支援開発のためのリグレッションテスト戦略。データベース依存なしのサンドボックスモード API テスト、自動化されたバグチェックワークフロー、同じモデルがコードを書いてレビューする AI のブラインドスポットを捕捉するパターン。

★ 262k89/100

Blackbox

NousResearch/hermes-agent

✓SAFE B

Delegate coding tasks to the Blackbox AI multi-model CLI.

★ 248k89/100

Openhands

NousResearch/hermes-agent

✓SAFE B

Delegate coding to OpenHands CLI (model-agnostic, LiteLLM).

★ 248k89/100

Plan Design Review

garrytan/gstack

!CAUTION B

Designer's eye plan review — interactive, like CEO and Eng review. (gstack)

★ 138k89/100

Plan Ceo Review

garrytan/gstack

✓SAFE B

CEO/founder-mode plan review. (gstack)

★ 138k89/100

Investigate

garrytan/gstack

✓SAFE B

Systematic debugging with root cause investigation. (gstack)

★ 138k89/100

Gstack

garrytan/gstack

✓SAFE B

Install and use gstack in Capy and Claude Code. Load at session start to check this machine, and for gstack planning, code review, debugging, QA, design, security, documentation, and shipping workflows.

★ 138k89/100

Document Release

garrytan/gstack

!CAUTION B

Release documentation audit. (gstack)

★ 137k89/100

Design Review

garrytan/gstack

!CAUTION B

Designer's eye QA: finds visual inconsistency, spacing issues, hierarchy problems, AI slop patterns, and slow interactions — then fixes them. (gstack)

★ 137k89/100

Design Consultation

garrytan/gstack

✓SAFE B

Design consultation: understands your product, researches the landscape, proposes a complete design system (aesthetic, typography, color, layout, spacing, motion), and generates font+color preview... (gstack)

★ 137k89/100

Cso

garrytan/gstack

✓SAFE B

Security audit: supported static findings; qualified profiles add reproduction and repair candidates. (gstack)

★ 137k89/100

Shadcn

shadcn/ui

✓SAFE B

Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working…

★ 126k89/100

Also for Codex

Questions about Codex skills

Are Codex skills safe?
474 of these 867 skills (55%) are SAFE, 161 CAUTION and 232 BLOCK. Skills that ship scripts can run code on your machine, so read the findings on any CAUTION or BLOCK page before you install.
What does "works in Codex" mean here?
The skill's own text mentions Codex. We have not tested the skill in Codex, so each page shows the claim, not a compatibility result.
How do I install a Codex skill?
427 of these skills document an install command in their repository; the commonest forms are git clone <url> (132 skills), npx skills add <owner/repo> (67 skills) and npm install (35 skills). The rest are installed from the repository itself. Each skill's page shows the commit our verdict covers, so you can pin to it.
Do these skills work in other agents?
725 of the 867 also mention at least one other agent in their text, such as Claude Code, Claude Desktop and Cursor. Each page lists every agent the skill names.

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.