Codex skills
Find curated Codex Skills with readable sources. Learn where Skills load, how to run or install them, and how to check your own Skill across Agents.
Agents
npx skilld run owner/repo/skillCodex is OpenAI's coding Agent for the terminal and the editor. It reads Agent Skills from a directory in your project or your home directory. skilld gives you a curated set of those Skills, each written by a person in their own repository, and one command that puts a Skill in front of Codex.
How Skills work in Codex
Codex loads a Skill from a SKILL.md file. Each Skill sits in its own directory.
- Project:
.agents/skills/<skill>/SKILL.md - Global:
~/.agents/skills/<skill>/SKILL.md
Amp and Zed read the same two directories. One install serves all three Agents.
Codex discovers Skills at startup. It keeps each name and description in context. When a task matches a description, Codex loads the full file and follows it.
To write your own, follow Write a Skill that works across Agents.
Start with a complete SKILL.md, then check discovery and task completion in Codex.
Install a Skill for Codex
npx skilld install owner/repo/skill --agent codexThe Skill lands in .agents/skills. Add --global to install to ~/.agents/skills instead. skilld detects the Agents in your project and on your machine. If it finds only one, omit --agent.
Start a new Codex session after an install. Codex reads the directory at startup.
Run a Skill without installing it
npx skilld run owner/repo/skillskilld run prints the Skill and writes no file. Give the command to Codex. It reads the Skill and follows it for this session. Nothing lands in your repository, and no lockfile changes.
Codex CLI Skills
The Codex CLI and the Codex editor extension share the .agents/skills directory in a project. A Skill you install once works in both.
Several lists on GitHub call themselves awesome Codex skills. They collect links. The registry admits fewer Skills, names the author of each, and links the exact source file. Read the file, then decide.
Curated Skills for Codex
Codex CLI work lives in the terminal next to the test runner, so these Skills cover testing and debugging. Skills here come from the curated registry; a person admitted each one.
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
Use when implementing any feature or bugfix, before writing implementation code
Use when determining what test coverage ALREADY exists for a specific change (a PR, Jira key, Tech Breakdown doc, Testmo CSV, changed paths, or named component). Triggers on "what's already tested", "does this PR have tests", "what coverage exists for", "is this component covered", or "which behaviors have no test today". This is a backward-looking inventory of existing coverage for a concrete change. Do NOT use it to recommend or decide which new tests to add ("should I add integration tests here", "are unit tests enough"), to design a test strategy or plan, to run or fix existing tests, or to explain testing concepts like the test pyramid or which layers a repo uses — those are all out of scope.
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
Confirms what a third-party library, remote API, build backend, or scraped document actually does before writing code that depends on it — throwaway probes that run in seconds, permissive clients that forward wrong arguments instead of rejecting them, per-endpoint docs that don't generalize, response shapes that make a "fast path" always-false, fakes that encode your assumption rather than the service's behavior, dry-runs that skip the step that fails, and large inputs that were silently read only in part. Use when integrating a new dependency or endpoint, writing a tolerated-status or error branch, choosing a client argument name, testing against a fake, or reviewing code that asserts an upstream contract.
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
The protocol behind every investigation skill. Use when AI research must proceed without you: search-plan gate, Fact/Inference/Assumption labels, confidence stacking, diffable outputs.
Compatibility pointer for the canonical postmortem skill.
All 98 Skills in Testing and debugging
Codex can also search the registry itself, with the skilld Skill or the skilld MCP server. See the setup
Read before you run. A Skill is instructions for your Agent. Open the source file first. What skilld verifies, and what it never claims .