Real numbers, no spin
Trust & Traffic Transparency
Every number below comes straight from our production database — no vanity metrics borrowed from other people's repos or npm packages. This page shows exactly how people use the directory and which AI systems and crawlers actually read it, updated live on a trailing 30-day window.
How people use AllMCPs
Who's reading the API (last 30 days)
Every request to our search, listing, and llms.txtendpoints is classified server-side from its User-Agent. We deliberately keep “AI Reads” limited to named AI assistants and their crawlers — generic web crawlers and unclassified agents are broken out separately below so the two never get blended into one inflated number.
AI assistants & their crawlers
33 hits · 1.4% of trafficLive chat/agent traffic (Claude, ChatGPT, Gemini, Perplexity, Cursor, Copilot, Windsurf) plus the offline crawlers those companies run to index content (ClaudeBot, GPTBot).
- ClaudeBot (Anthropic crawler)13 · 0.6%
- Gemini / Google10 · 0.4%
- ChatGPT / OpenAI9 · 0.4%
- GPTBot (OpenAI crawler)1 · <0.1%
Other crawlers & bots
2,158 hits · 93.2% of trafficGeneral-purpose web crawlers, link-preview bots, and unnamed automated clients — not AI systems, but still worth counting honestly.
- Meta / Facebook bot999 · 43.2%
- Other bots & crawlers904 · 39.0%
- Custom agents178 · 7.7%
- Amazonbot77 · 3.3%
Human visitors (browser requests to the API)
124 hits · 5.4% of trafficReal browsers hitting API endpoints directly, separate from normal page views.
- Browser (Human)124 · 5.4%
Why we're open about this
Our robots.txt explicitly allowlists AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Amazonbot, Bytespider) to read our catalog, search API, and llms-full.txt. We set ai-input=yes so answer engines can cite us, and ai-train=noso the catalog isn't used as training data. Nothing on this page is gated behind a login or gamed by us — it's the same data our own team looks at.