Free · public pages only · a few minutes

What is your website teaching the machines?

Campus chatbots, RAG systems, and the public LLMs all read your website as the truth. When two of your pages disagree, the model does not flag it. It picks one and states it with confidence. The scan reads about 30 of your public pages, pulls out every fact that can be checked, and shows you where the site contradicts itself, where it is out of date, and where the alt text or meta description says one thing while the page says another.

Run the scan

What we check

Every fact a chatbot might repeat

  • Deadlines and dates Application, aid, and registration deadlines that differ from page to page, or have already passed.
  • Dollar figures and counts Tuition, fees, enrollment, class sizes, and any number that shows up twice with two values.
  • Requirements Test policies, credit requirements, and eligibility rules that one page relaxed and another did not.
  • Contact details Phone numbers, emails, office names, and addresses that moved on one page and not the others.
  • Events Events still advertised after they happened, and the same event listed on two different dates.
  • Meta descriptions The summary search engines and chatbots read first, checked against what the page actually says.
  • Alt text Image descriptions that state a fact the visible page contradicts or no longer supports.

How it works

Three steps, a few minutes

  1. Give us the site

    Your name, your organization, and the homepage address. No account needed to start.

  2. We read about 30 pages and pull out the facts

    We follow your own links, favoring admissions, costs, aid, events, and contact pages, and extract every date, figure, requirement, and contact detail with the sentence it came from.

  3. One pass finds the contradictions

    All of those facts are compared at once. You sign in once, with Google or your email, to open the result. Every finding cites the page and the exact words.

Run the scan

All fields are required unless marked optional.

Sign in with this email, or with Google on the same address, to open the result.

Your homepage is enough. We stay on the same site and read up to 30 public pages from there.

Public pages only. We never log in to anything. Nothing is stored beyond the facts we extract and the result.