Home / About this site

About this site

What this site is, where the data comes from, and how we judge usability.

What this is

An index of AI tools and MCP servers. We do not build the tools. We do three things: collect entries scattered across GitHub, SkillHub, SkillsMP and several curated lists; label each one with whether it is usable (Chinese UI / ICP filing / a free tier); and send you straight to the tool's own site with no interstitial page.

The catalogue currently holds 7,555 skills and MCP servers and 1,501 AI tools; the sitemap lists 15,595 pages.

Where the data comes from

Counts by source, computed from the database when this page was generated:

  • awesome:punkpeye/awesome-mcp-servers — 4,255
  • github — 969
  • skillhub — 806
  • awesome:abordage/awesome-mcp — 630
  • awesome:steven2358/awesome-generative-ai — 330
  • skillsmp — 174
  • awesome:e2b-dev/awesome-ai-agents — 157
  • awesome:rohitg00/awesome-devops-mcp-servers — 115

Each source has its own page on this site: GitHub · SkillHub · SkillsMP · curated awesome lists. Every change to the site is recorded in the changelog.

Sources starting awesome: are curated lists. 6,102 entries take their description from the item's own repository README.

How we judge "usable"

Three signals, each with a stated test you can re-run yourself:

  • Chinese UI (870) — from the language of the tool's own page.
  • ICP filing (283) — we record the filing number itself, which you can check on the MIIT registry.
  • Free tier (535) — the pricing or landing page explicitly states a free plan or trial. A card-required trial does not count, one-off credits do not count, limited-time promotions do not count.

Where we have no evidence we write n/a rather than infer. That is why some rows show n/a for price: not that the tool is paid, but that we found no evidence.

How we keep it fresh

  • Daily harvest from GitHub / SkillHub / SkillsMP / curated lists.
  • Weekly link health checks on official sites and GitHub repositories; a repository that no longer exists is labelled as such and its link becomes non-clickable.
  • Numbers on pages are computed from the database at build time, never cached.

Honest disclosure

We are an automated aggregator, not a review outlet. No human scores, no editor's picks, no paid placement. Rankings come from public numbers (for example GitHub stars) that you can open and compare yourself.

If something is wrong — a price changed, a site moved, a tool shut down — write to us and we will fix it.

Contact

Email: david@anclum.com

← Back to index