Vantage by URRU

How this works

  1. You paste a URL. We fetch that page over plain HTTP, identifying ourselves as VantageBot/1.0.
  2. We look for robots.txt, sitemap.xml and llms.txt, then pick up to four more pages that look like pricing, services, about and contact.
  3. We extract the text. We never run JavaScript. What we extract is what a non-rendering AI crawler receives.
  4. That text goes to Claude Sonnet 5 with no other context — no brand name, no hints, and an explicit instruction to ignore anything it might recognise from training data.
  5. The model returns what it understood and what it could not determine. It never returns a number: the score is computed here, from its findings and from what we measured while fetching.

About VantageBot

VantageBot only fetches a site when a person pastes that URL into vantage.urru.ai. It reads at most five pages, obeys an 8-second timeout per page, and never submits forms or follows anything but ordinary links. It does not crawl on a schedule.

To keep it out, add this to your robots.txt:

User-agent: VantageBot
Disallow: /

Questions: office@urru.ai