Free tool · Instant · Nothing stored

What does an AI assistant actually see on your site?

When someone asks ChatGPT or Perplexity for a business like yours, the assistant fetches pages and reads the raw HTML. This checks whether yours tells it who you are, what you do, and how to reach you — or whether it comes back with almost nothing.

Takes about a second. No signup, no email, and we don't store the address you enter.

What it looks at

Seven things a machine needs to read you.

Every one of these is checkable from the outside in under a second, which is why this is a tool and not a proposal.

Readable without JavaScript

The big one. If your page is assembled in the browser, an assistant reading the raw HTML may see an empty shell. This counts the words actually present.

Structured data

Whether there's a schema block labelling which text is your name, address, and hours — the difference between a machine knowing facts and guessing at them.

The facts people ask for

Name, contact, address, hours: machine-readable, present only as prose, or absent. Contact accepts a phone or an email; address and hours are skipped for sites that aren't walk-in businesses.

AI crawler access

Whether robots.txt blocks GPTBot, ClaudeBot, PerplexityBot, Google-Extended and friends. Blocking them is a fair choice — it should just be a deliberate one.

Title & description

The one-line summary every search engine and assistant quotes when it mentions you. Missing means something gets guessed on your behalf.

llms.txt & structure

The emerging convention for telling assistants what your site is, plus whether the page has one clear heading and a canonical address.

The capability behind this work