Skip to content

Claude Skills

Claude Skills Directory

Find Claude skills that are easier to trust before you install them. Skills Directory indexes reusable agent skills and highlights security grades, source attribution, install paths, and practical use cases.

Browse Claude skillsLearn what skills are

Indexed skills

592,677

Search focus

Claude skill

Security context

Scanned

Top matching skills

What are Claude skills?

Claude skills are reusable instruction packages, usually centered on a SKILL.md file, that teach Claude a workflow, domain, style, or tool pattern. They are useful when you want Claude to repeat a process reliably without re-explaining it every session.

Why use a security-scanned directory?

Skills can include instructions, scripts, references, and install commands. A directory with security grades helps you spot prompt injection, credential handling, shell execution, network access, and other risk signals before you copy a skill into your agent environment.

How to pick the right skill

Start with your workflow, check the description and source, review the security grade, then install only the skills you actually need. Favor narrow skills that do one job well over broad bundles that try to control every agent behavior.

Why security-scanned skills matter

Skills can influence agent behavior, suggest commands, include helper files, or touch code and data. Skills Directory adds security context so you can inspect risk before installing a workflow from a public repo.

See the security model

Frequently asked questions

Are Claude skills the same as MCP servers?
No. A skill teaches Claude how to perform a workflow. An MCP server gives Claude access to external tools or data. Many advanced setups use both: MCP for capability, skills for process.
Are all skills on Skills Directory free?
Skills Directory links to community and open-source skills. Individual upstream projects may have their own licenses, but browsing and discovery on Skills Directory is free.
What makes a Claude skill safe?
A safer skill is narrow, transparent, source-attributed, recently maintained, and free of suspicious instructions such as credential exfiltration, hidden network calls, or unreviewed shell execution.