Show HN: I built a search engine for llms.txt sites

3 points

a month ago

More and more developer tools are adopting the llms.txt standard to build AI-friendly versions of their docs. The problem is that it's very hard to search across them. So I crawled millions of webpages to build Statespace, a search engine for llms.txt sites. And it's free, no API keys required.

You can run plain queries to search across all sites:

  mcp server setup
  vector database embeddings
  oauth2 refresh token

Or scope your queries to a specific site with "site: query"

  stripe: webhook verification
  mistral.ai: function calling
  docs.supabase.com: edge functions auth

Quotes work like Google for exact phrases:

  "context window limit"
  vector database "semantic search"
  stripe: "webhook signature verification"

This is still a work in progress, so looking for feedback and beta testers!

- Search for humans (website): https://statespace.com/

- Search for agents (CLI, SDK, MCP, skill): https://github.com/statespace-tech/statespace