Claude Code Skills·Claude Skills·The open SKILL.md registry for Claude
ClaudSkills › Stats

ClaudSkills by the numbers

An open registry of Claude Code Skills — community-contributed SKILL.md files for Anthropic's Claude Code agent. Updated twice daily from 26 public sources. No author pays for inclusion or ranking.

Last refreshed 2026-05-23 · Refreshes at 02:00 + 14:00 UTC daily · Free, no ads, no tracking

69,000+ SKILL.md files in the catalog
5,000+ skill authors credited
2,400+ author profile pages
1,500+ source repositories mined
26 public mining sources
10 broad categories
118 sub-categories
58 orthogonal tags

For the full experience including multi-signal Quality Score and one-click install features for each skill — upgrade to Pro.

How the catalog is built

The promise: ClaudSkills is an objective discovery layer for Claude Code Skills. No skill author pays anything — no listing fee, no acceleration, no ranking boost, no featured placement, no sponsored Pro Collections, no "verified author" upsell, no affiliate kickback from external directories. The catalog is mined automatically; admission is rule-based; ranking (when Pro is enabled) is derived from content signals only.

Category breakdown

CategorySkillsBrowse
General36,000+/category/general/
Engineering13,000+/category/engineering/
Security3,500+/category/security/
Dev Tools3,000+/category/tools/
Content3,000+/category/content/
Science & Research2,500+/category/science/
Product2,500+/category/product/
Growth1,500+/category/growth/
Sales1,200+/category/sales/
Advertising200+/category/ads/

Counts are rounded to the nearest hundred so the page stays stable between miner cycles. Live counts are derived from /data/skills.json on every refresh.

The 26 mining sources

Sources span code hosts (GitHub, GitLab, Codeberg, HuggingFace), aggregators (awesome lists, marketplaces, RSS feeds), and discussion surfaces (Reddit, HackerNews, dev.to, Mastodon, Bluesky, Telegram, YouTube). Each source is rate-limited and wrapped in a fail-soft section so one broken endpoint cannot kill the run.

  1. Netlify Forms — user submissions
  2. GitHub code search (101 query variants, sort-window splitting)
  3. Awesome lists — parses GitHub READMEs
  4. GitHub Topics (31 topics, auto-growing)
  5. DuckDuckGo web search (LLM query expansion auto-grows queries)
  6. Standalone SKILL.md files — gists + raw markdown
  7. GitLab
  8. Codeberg (Gitea)
  9. HuggingFace
  10. Reddit JSON API (kept for future OAuth)
  11. HackerNews Algolia search
  12. Bluesky public AppView
  13. Mastodon (11 instances)
  14. dev.to API
  15. YouTube search-results scrape + video-page follow-through
  16. Wayback Machine CDX (finds renamed/deleted repos)
  17. Stargazer mining (graph expansion from seed repos)
  18. Author repo enumeration (existing skill authors' other repos)
  19. Topic co-occurrence (grows discovered topics for next run)
  20. Telegram public channels
  21. VSCode + Open VSX marketplaces
  22. Catalog README mining (curated lists)
  23. Newsletter RSS (Latent Space, AI News, etc.)
  24. Brave Search API (when key available)
  25. LLM query expansion (Anthropic API — generates web-search queries for next run)
  26. GitHub Gists active search

Full source code at scripts/skill_miner.py. Every source has its own admission counter, hash-deduplication against prior runs, and per-source budget cap.

Direct data access

Catalog JSON

Full catalog as a single JSON file. One object per skill — slug, name, description, category, sub-category, tags. Refreshed every miner cycle. CORS-open for client-side use.

GET https://claudskills.com/data/skills.json

Per-skill SKILL.md

Individual SKILL.md markdown files served via Cloudflare R2 for read-only mirroring at scale. Identical to the version installed when a developer runs claudskills install <slug>.

GET https://claudskills.com/skills/<slug>/SKILL.md

XML sitemaps

Three sub-sitemaps referenced from the index — pages, skills, authors. Standard <urlset> format with lastmod and priority.

GET https://claudskills.com/sitemap.xml

RSS / JSON / Atom feed Coming Phase 2

Last-100-skills feed in three formats served via a Cloudflare Worker, with WebSub hub announcement for low-latency aggregator discovery. See the promotion-engine action plan for status.

HuggingFace dataset mirror Coming Phase 2

Full catalog mirrored as a HuggingFace dataset (Parquet + JSON), refreshed daily. Tagged for LLM training pipeline discovery.

Public GitHub catalog mirror Coming Phase 2

Read-only repository at github.com/claudskills/catalog-public. Same data as the JSON above, with per-row attribution + a CITATION.cff for academic citation.

Quality bands (Pro)

Free browsing exposes the full unranked catalog. Pro subscribers see a multi-signal QualityScore on every skill in the desktop app, computed server-side from content structure and frontmatter depth. The score is rescaled to a stable [50, 100] reference band so adding new skills never reshuffles existing scores. See the about page for the methodology in full.

Citation

If you reference ClaudSkills data in a paper, blog post, or library: please cite ClaudSkills with the URL https://claudskills.com/stats/ and the access date. A CITATION.cff entry will land alongside the public catalog mirror once Phase 2 of the promotion engine ships.

Browse the full catalog →