Agent Skills
Find, install, and run the best agent skills.
An agent skill is a folder with a SKILL.md file that teaches an agent how to do something new. Every agent implements it differently — where skills live, how you install them, what’s worth trusting. These guides cover one ecosystem at a time: exactly how each system works, and which skills are worth your time.
- Method
- Official docs and repository source, read directly. Never paraphrased search results.
- Independence
- No vendor affiliation, no sponsored placement, no invented authors.
- Scope
- Every ecosystem covered in full, never stretched thin.
Discover
Find a skill, browse a topic, or ask for one.
82 real, source-verified skills — searchable by name, task, tool, creator, or source, not just titles. 35 have a full decision-support profile.
- Know the area?
- Browse 9 topics →
- Know only the outcome?
- Ask Doctor →
Ecosystems
Every agent, covered in full.
A cluster ships when the research is done. Each one is built from that ecosystem’s own documentation and source, so the paths and commands are its own rather than a neighbour’s.
Hermes Agent, in four parts.
Live cluster
Nous Research's open-source personal AI agent — what its skills are, the best ones to install, and exactly how to set them up.
- 01Hermes Agent Skills: The Practical GuideWhat Hermes Agent skills are, where they live, and the eight sources you can install them from.
- 02The Best Hermes Agent SkillsEight Hermes skills worth installing, and exactly what each one is good for.
- 03How to Install Hermes Agent SkillsInstall Hermes Agent and start adding skills — every command, verified.
- 04Hermes Agent Skills on GitHubWhere to find Hermes skills on GitHub, and how to tell a good repo from a risky one.
Claude Code, in four parts.
Live cluster
Anthropic's agentic coding CLI, and the reference implementation of the open Agent Skills standard — what its skills are, the best ones to install, and exactly how to set them up.
- 01Claude Code Skills: The Practical GuideWhat Claude Code skills are, where they live, and the frontmatter fields that only work in Claude Code.
- 02The Best Claude Code SkillsThe Claude Code skills worth installing, verified against the actual repos, not another roundup.
- 03How to Install Claude Code SkillsThe three real ways to add a Claude Code skill, and when to use each one.
- 04Claude Code Skills on GitHubWhere to find Claude Code skills on GitHub, and how to vet one before you run it.
Codex, in five parts.
Live cluster
OpenAI's terminal coding agent — what its skills are, which directory they actually load from, and which ones are worth installing.
- 01Codex Skills: The Practical GuideWhat Codex skills are, the two skill directories that both still work, and how Codex decides to load one.
- 02The Best Codex SkillsThe Codex skills worth installing, taken from the real catalog rather than another roundup.
- 03How to Install Codex SkillsThree ways to add a Codex skill, and which directory each one actually writes to.
- 04How to Create a Codex SkillWrite a SKILL.md that Codex will actually trigger, then package it so other people can use it.
- 05Codex Skills vs AGENTS.mdWhich one to reach for, and why context cost is the deciding factor.
Cursor, in five parts.
Live cluster
The AI code editor — where its skills load from, including the Claude Code and Codex directories it reads for free, and how skills differ from the rules you already have.
- 01Cursor Skills: The Practical GuideAll eight directories Cursor loads skills from, and the two frontmatter fields that only exist here.
- 02Cursor Skills vs RulesCursor's own migration tool defines which rules should have been skills. Here's what it converts, and what it leaves alone.
- 03How to Install Cursor SkillsFive ways to get a skill into Cursor, and the one that needs no work at all if you already run Claude Code.
- 04The Best Cursor SkillsThe 19 already installed, the reviewed marketplace, and the test for whether an outside skill will work here.
- 05Cursor Plugins: What They Are and What's Inside ThemAn inspected catalog of the official marketplace plugins — every skill, subagent, and rule each one ships, enumerated.
OpenClaw, in five parts.
Live cluster
The open-source personal assistant with the largest skill registry of any agent covered here — six load paths, real dependency gating, and a trust model worth understanding before you install anything.
- 01OpenClaw Skills: The Practical GuideThe six directories OpenClaw loads from in priority order, and the gating metadata that decides what the agent ever sees.
- 02How to Install OpenClaw SkillsEvery install form and the directory it actually writes to — including two commands that look alike and aren't.
- 03The Best OpenClaw SkillsThe 51 skills already in the box, and why ClawHub's download counts can't rank the rest.
- 04Are OpenClaw Skills Safe?What ClawHub's audits check, what they can't catch, and what the independent research actually found.
- 05How to Create an OpenClaw SkillWrite a skill that hides itself when its dependency is missing, then publish it to ClawHub.
The format
What counts as an agent skill.
A skill is a folder with a SKILL.md file — metadata plus instructions telling an agent how to perform a specific task — and, optionally, bundled scripts, references, and templates.
Agents load skills through progressive disclosure: only a name and description at startup, full instructions when a task actually matches, and any bundled files only when the instructions call for them. That keeps a large skill library cheap to keep on hand.
The format was originally developed at Anthropic and released as an open standard, which is why it now shows up across a wide range of agents rather than staying locked to one product. What differs between agents is everything around the format: where skills live, how they’re installed, what trust and security model applies, and whether the agent can write its own skills.
That’s exactly what trips people up moving between ecosystems: a command from a Claude Code tutorial doesn’t map onto Hermes’s install flow, “official” doesn’t mean the same thing on every hub, and Codex reads skills from a directory its own source marks deprecated. The guides here cover each ecosystem specifically — where its skills actually live, which ones are worth your time, and how to tell a well-built one from a risky one before you run it.
Across ecosystems
Questions that aren’t about one agent.
The format itself, where skills come from, what travels between agents, and why a skill that looks correct sometimes does nothing at all.
- 01Agent Skills Not Working: A Diagnostic ReferenceWhy a skill doesn't load, doesn't appear, or doesn't trigger — sorted by mechanism, traced to the issue or document it came from.
- 02Where to Find Agent SkillsSeven kinds of source, what each one's provenance is worth, and what their download numbers actually count.
- 03SKILL.md: The Agent Skill File FormatEvery frontmatter field with its real constraints, and where five runtime implementations diverge from the spec.
- 04Agent Skills Compatibility: Which Agents Support SKILL.mdWhat support means in each runtime, with explicit evidence levels — and the field that silently breaks a skill.
- 05Agent Skills vs MCPTwo layers, not two options — argued from the manifests and hooks that already ship, rather than from analogy.