VoltAgent/awesome-agent-skills

What it is — awesome-agent-skills is a curated collection of Agent Skills: small, focused bundles of instructions and tooling that extend what an AI coding assistant can do. If a code assistant is like a worker, a Skill is a specialty manual — e.g., "how to write Cypress tests correctly" or "how to create Word documents."

Unlike many similar repositories that are bulk-generated by AI, this one is explicitly hand-picked by humans — the README repeats this point several times. It's maintained by VoltAgent together with the community, and claims to be the "most contributed" Agent Skills repository.

Scale and scope — The badge says 1,497+ skills. They come from official teams (Anthropic, Google, Stripe, Microsoft, OpenAI, Figma, MongoDB, Cloudflare, Vercel, Supabase, and many more) plus community submissions. The README has a long table of contents listing ~60+ contributing teams.

Compatibility — The skills work across popular coding tools: Claude Code, Codex, Antigravity, Gemini CLI, Cursor, GitHub Copilot, OpenCode, and Windsurf. The README says paths and documentation are in the linked tables.

What's inside (a sample) — Examples pulled directly from the README:

• Claude official skills: create/edit Word docs, PowerPoint, Excel, PDFs, generative art, frontend design, Slack GIFs, web app testing, brand guidelines.
• VoltAgent's own skills: project setup, best practices, core framework reference, docs lookup for their TypeScript framework.
• SerpApi: web search across 130+ engines (Google, Scholar, Maps, Flights, Hotels, Shopping).
• Crawlbase: crawl any URL as raw HTML, clean Markdown, or screenshot; manage stored pages.
• TestMu AI: test-automation skills for ~30 frameworks (Selenium, Playwright, Cypress, Jest, pytest, etc.).

Sponsors — Three sponsors appear at the top: TestMu AI (an AI-native testing cloud), Crawlbase (web data infrastructure), and SerpApi (a web search API). Each also contributes its own skills listed in the collection.

Format — The README organizes everything with collapsible sections per team. Links use officialskills.sh URLs and GitHub links. The repo also carries an "Awesome" badge (from the awesome.re curated-list directory) and has an associated Discord server and a separate repo for official MCP servers.

Bottom line — If you use an AI coding assistant and want it to do specialized work (tests, docs, web access, design) with real, vetted playbooks rather than generic AI guesses, this is a central library for plugging those capabilities in.