Home › Learn › What is a Claude Skill?
What is a Claude Skill?
A Claude skill is a SKILL.md file stored under ~/.claude/skills/<skill-name>/ that tells the Claude Code CLI agent how to perform a specific, named task. When Claude Code starts a session, it auto-discovers every SKILL.md in that directory and invokes the right one based on
A Claude skill is a SKILL.md file stored under ~/.claude/skills/<skill-name>/ that tells the Claude Code CLI agent how to perform a specific, named task. When Claude Code starts a session, it auto-discovers every SKILL.md in that directory and invokes the right one based on the task you ask it to do.
Each SKILL.md has YAML frontmatter (name, description, optional model, allowed-tools, hooks, user-invokable) followed by Markdown instructions. The frontmatter is the metadata Claude uses to decide when to invoke the skill; the body is what to do once invoked.
Claude skills are also called Claude Code skills (more precise, since they're scoped to the Claude Code CLI) and occasionally just skills in community shorthand. The terms are interchangeable.
Why skills exist: the alternative is pasting the same multi-step instructions into Claude every time you want a particular task done. Skills package that boilerplate into a discoverable, version-controlled, shareable file. The ClaudSkills catalog indexes 67,000+ open skills from public GitHub repositories, free to install.
Top 24 indexed skills
- divorce-ct-stephane-boghossianGeneral
Claude skill that turns Claude into a Connecticut-specific divorce planner — nine operating modes from pre-flight intake to post-judgment modification, modeled on Untangle.us's feature surface and grounded in…
- create-paid-skillEngineering
Use this skill when the user asks to "create a paid skill", "monetize an API", "create a SKILL.md", "set up x402 payments", "make a paid API for agents", "create…
- cognitive-load-theoryGeneral
Sweller's Cognitive Load Theory (CLT) for agents writing skill content, designing prompts, building UI/dashboards, and authoring documentation. Working memory holds roughly 4 chunks at a time; CLT classifies…
- build-help-skillGeneral
Use when the user wants to "scaffold a help skill", "add a /<plugin>:help command", or "build a plugin index skill", or wants to give a plugin an orientation surface that lists its…
- validate-pluginGeneral
Validate a Claude Code plugin directory against the official Anthropic spec and Intent Solutions enterprise standard. Runs structural validation (plugin.json fields, file references, permissions) and content…
- book-to-skillGeneral
Converts a technical book (PDF or EPUB) into a structured Claude Code skill — extracting frameworks, mental models, principles, techniques, and anti-patterns the author crystallized. Use when the user wants to study…
- check-skill-pairGeneral
Audits pair-level integrity of a primitive-pair (the artifact `/build:build-skill-pair` produces) by walking the four required artifact slots — principles doc, `build-<primitive>/SKILL.md`,…
- skill-checkGeneral
Validate Claude Code skills against the agentskills specification. Catches structural, semantic, and naming issues before users do.
- skill-infrastructureEngineering
Use when designing the deterministic health-tooling layer for a skill library, diagnosing why an existing library is decaying invisibly, deciding which categories of automated check to add (inventory, protocol…
- skills-curator-liteEngineering
The intelligence layer for Claude skills — without Python. Maintains a trust-rated catalog (curated entries + live GitHub topic search via curl), identifies what fits your project (stack, deps, CLAUDE.md), recommends…
- check-help-skillGeneral
Use when the user wants to "audit a help skill", "review my plugin index", or "verify my help-skill is up to date". Audits a plugins/<plugin>/skills/help/SKILL.md against the…
- skill-forgeScience
Generate a production-ready AbsolutelySkilled skill from any source: GitHub repos, documentation URLs, or domain topics (marketing, sales, TypeScript, etc.). Triggers on /skill-forge, "create a skill for…
- skill-scoreTools
One repeatable pass that keeps the skill pack itself effective — the one surface no other scorecard grades. Runs the skill-effectiveness scorecard (`fak skill-effectiveness-scorecard`) over every…
- continuous-learningGeneral
Continuous learning system that monitors all user requests and interactions to identify learning opportunities. Active during: (1) Every user request and task, (2) All coding sessions and problem-solving activities,…
- claude-skill-spec-auditGeneral
Audit skill SKILL.md files for compliance with the agentskills.io specification and house conventions. Checks frontmatter fields (name, description, compatibility, metadata, argument-hint), metadata sub-fields…
- claudeceptionEngineering
Claudeception is a continuous learning system that extracts reusable knowledge from work sessions. Triggers: (1) /claudeception command to review session learnings, (2) "save this as a skill" or…
- claudskills-installGeneral
Install Claude Code skills from claudskills.com on the user's request. Fetches SKILL.md by slug from the catalog and writes it to ~/.claude/skills/<slug>/SKILL.md. Pairs with Claude Code Channels…
- extract-skillGeneral
(What) Format a SUCCESS WP or unit-contract into a TPMN SKILL.md and store in .claude/skills/. (When) After /archive-work SUCCESS, or anytime human says "extract this as a skill" or "make this…
- lythoskill-curatorScience
Skill discovery engine. Scans your local cold pool (~/.agents/skill-repos), indexes all SKILL.md frontmatter into REGISTRY.json + catalog.db, and lets you query by type, niche, source, or keyword. The CLI is a dumb…
- search-skillGeneral
(What) Search .claude/skills/ and .gem-squared/external-skills/ for skills and skill directories relevant to the current task. (When) During /plan-work when domain capability is needed, or anytime human asks…
- agent-experienceSecurity
Audit the developer experience of a product, SDK, docs site, or SKILL.md by dropping multiple Claude subagents at it with only a tiny task prompt and real tools (WebFetch, Bash, Write). Agents must discover the docs…
- promote-skillGeneral
Use when publishing a SKILL.md-style agent skill across uGig, sh1pt, GitHub/gists, and follow-on skill marketplaces such as ClawHub, Goose, LobeHub, Kilo, Skillstore, FreeMyGent, ClawMart, Manus, VS Code Agent…
- skillsmithGeneral
Discover, install, compare, and manage Claude Code skills. Search the registry, get recommendations, validate skill quality, and manage your installed skills.
- aif-skill-generatorTools
Generate professional Agent Skills for AI agents. Creates complete skill packages with SKILL.md, references, scripts, and templates. Use when creating new skills, generating custom slash commands, or building…
Frequently asked questions
Where do Claude skills live on disk?
~/.claude/skills/<skill-name>/SKILL.md on macOS and Linux, or the equivalent %USERPROFILE%\.claude\skills\<skill-name>\SKILL.md on Windows. The directory name (slug) is what Claude uses to reference the skill internally. Some skills also include helper scripts, sample data, or sub-skills in additional files within that directory — only SKILL.md is required.How does Claude decide when to use a skill?
description: field in each skill's frontmatter at session start. When you give Claude a task, it picks the skill (if any) whose description best matches what you asked. You can also invoke a skill explicitly by name (e.g. "use the code-review skill on this PR"). For finer control, the frontmatter can declare user-invokable: true so the skill only fires when named, never auto-selected.Are Claude skills the same as Anthropic Skills (the platform feature)?
How do I install a Claude skill?
mkdir -p ~/.claude/skills/<slug>, (2) curl -fsSL https://claudskills.com/skills/<slug>/SKILL.md -o ~/.claude/skills/<slug>/SKILL.md, (3) restart Claude Code, (4) run /skills to confirm. Or use the ClaudSkills desktop app for one-click install on macOS / Windows.