📥 URL Importer

Import from URL

Paste any article URL. The server scrapes it and extracts every fact scoring 7+ — not just one — then saves them to the drafts queue for Bulk Review.

💡 Rich sources (Wikipedia, long articles) yield the most facts. Extraction takes ~15–40s.

How it works

  1. Server fetches the URL via Firecrawl (full article, up to 18k chars)
  2. Claude extracts every distinct fact scoring 7+, ignoring the obvious
  3. Each fact is geocoded and inserted into fact_drafts
  4. You review and approve in Bulk Review