learn-skills.dev
Curated high-quality AI Agent Skills. Search, install, copy and share. Works with Claude Code, Cursor, OpenClaw, and other AI coding tools.
Learn Skills Dev catalogs installable AI agent skills
A web app and crawler that aggregates AI agent skills from community leaderboards and serves them through search, install, and share flows.
What the site does
Learn Skills Dev is a curated directory of AI agent skills that works with Claude Code, Cursor, OpenClaw, and similar coding tools. The web app lets users search, install, copy, and share skills, and the README is translated into many languages including Chinese, Japanese, Korean, French, German, Spanish, Italian, Russian, and Arabic. The goal is to make community produced skills discoverable in one place rather than scattered across separate repositories.
Where the data comes from
The crawler currently pulls from skills.sh, a community curated skills leaderboard that ranks by total installs, recent trending growth, and daily hot installs. Planned providers include GitHub Trending and curated awesome lists for agent skills. Skills not tracked by any provider can be added manually through a data/manual_skills.json file, which the crawler fetches the SKILL.md for and keeps across runs, deduplicating against later provider data by highest all time installs.
Outputs and usage
The crawler writes data/skills.json with the three leaderboards, data/skills_index.json for the website with descriptions and deduplication by id, and data/feed.json holding the top 50 entries from each board. Cached SKILL.md files are stored under data/skills-md, and an RSS feed publishes meaningful changes to avoid spamming subscribers. Local development uses bun install followed by bun run crawl. The repository shows 200 stars and 34 forks as of August 2026.
Editorial conclusion
Learn Skills Dev records 200 stars and 34 forks as of the August 2026 snapshot. It aggregates agent skills from community leaderboards and serves them through a multilingual web app with CSV, JSON, and RSS outputs.
Community notes