crawlberg turns websites into cleaner RAG-ready text

crawlberg is a Rust tool for crawling websites and extracting useful content. It checks robots rules and sitemaps, and it can use when a page needs JavaScript before the content appears. Its output is cleaned Markdown plus structured details such as links, metadata, JSON-LD, and Open Graph data.

This gives a RAG system text that is easier to split into chunks, instead of messy raw HTML. BM25 filtering lets the crawler follow pages related to a search query instead of pulling an entire site. SSRF protection is on by default, so user-submitted URLs are blocked from reaching private networks, local addresses, or cloud metadata endpoints.

It , uses an , and avoids dependence on one vendor.

Key points

  • crawlberg is an Rust crawl-and-scrape engine.
  • It supports robots rules, sitemaps, and JavaScript-rendered pages through .
  • It outputs cleaned Markdown and for .
  • BM25 filtering can reduce irrelevant crawling.
  • SSRF protection and are built in.

Sources covering this story (2)

Read original