Skip to content
← Tools / Skills
Skill · .md packageAnthropic/10 trustVerified

Anthropic Agent Skills

Anthropic's official Agent Skills repo — document skills, frontend design, webapp testing, and the Agent Skills spec itself.

Visit official site →

About

The reference implementation of Anthropic's Agent Skills format, installable as a Claude Code plugin marketplace or via the Skills CLI. Ships self-contained skills including the document skills that power Claude's docx/pdf/pptx/xlsx editing, plus example skills for frontend design, webapp testing, MCP building, and creative work. Also hosts the open Agent Skills spec and a skill template.

Note: most skills are Apache 2.0, but the document skills (docx/pdf/pptx/xlsx) are explicitly source-available, not open source, per the README.

What it can do

  • Document creation/editing for docx, pdf, pptx, xlsx (the same skills behind Claude's document features)
  • Frontend-design skill (non-templated visual design decisions)
  • Webapp-testing skill (Playwright-based web app testing)
  • mcp-builder + skill-creator (meta-skills for building MCP servers and new skills)
  • Creative skills (algorithmic art, canvas design, theme factory)
  • The open Agent Skills spec itself, for authoring your own

Setup

  1. Claude Code: /plugin marketplace add anthropics/skills → /plugin install document-skills@anthropic-agent-skills (or example-skills@anthropic-agent-skills)
  2. Any agent: npx skills add anthropics/skills --skill <name>
  3. Also built into Claude.ai (paid plans) and the Claude API (Skills API).

Prerequisites

A compatible agent (Claude Code, Claude.ai, or API access with Skills beta).

Use cases

codedesignautomationtestingmedia

Works with

Claude CodeAny agent (.md)

Trust score — 9/10

/10 trustVerified

Official Anthropic repo, the most-starred skills repo from a model vendor (180k), excellent docs and a public spec. Minor caveat: most skills are Apache 2.0 but the document skills are explicitly source-available, not open source, per the README. No data-access concerns beyond the agent's own tools.

★ 180,708 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
9.6/10
Maintenance · recency, releases, issue load
7.6/10
Openness · license + docs
8.6/10
Manual review · sanity read + red-flag check
9.0/10
  • · 180,708 GitHub stars (log-scaled vs catalog peers)
  • · last push 3d ago; 0 releases/12mo; 1401 open issues
  • · manual review: authoritative source; doc-skills restrict commercial use

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.