bisicaDocs FREE CHATGPT CHECK

Can AI read your site: what the site check reads

The free site check reads a website’s robots.txt and homepage. It shows which of 10 AI crawlers are told to stay out, what the homepage sends before any script runs and whether the site has an llms.txt file and a plain-text copy. It asks no AI tool and gives no score.

Updated · 2 min read

On this page

Run it

  1. Open Can AI read your site.
  2. Type a website, for example yourcompany.com.
  3. Press Check the site. It takes a few seconds.

Four requests

The check makes four requests to the site under bisica’s own name:

Request What it shows
Your robots.txt What each of 10 crawlers from OpenAI, Anthropic, Perplexity, Google and Microsoft is told
Your homepage The number of words before any script runs, the page’s size, a phone number written as text, the schema.org labels it carries and any tag that limits search engines (noindex, nosnippet, nocache, noarchive)
Your homepage as plain text Whether the site sends a plain-text copy to an AI agent that asks for one
Your llms.txt Whether the file is there

It also says whether the site is served through Cloudflare, since Cloudflare has its own settings for AI crawlers.

Reading the crawler list

Crawlers are shown in two groups.

  • Crawlers that read pages for answers. If one of these is told to stay out, that AI tool’s search can’t read the page we checked.
  • Crawlers that collect training data. Keeping these out is your choice and doesn’t stop the answer crawlers.

Google-Extended is a control name with no crawler of its own. The check records it and leaves it out of the list. AI crawlers and robots.txt explains each crawler.

Firewalls and other pages

  • A firewall or bot rule can still turn a crawler away, which this check can’t see. The free snapshot tests for it by asking for your pages under six AI crawlers’ names.
  • It reads the homepage only. Other pages can have other tags.
  • It asks no AI tool. To see who ChatGPT mentions, run the free ChatGPT check.

Kept for one hour

We keep the website you type and what we read from it for one hour. See How we handle your data.

Run it from a chat

The same check runs inside Claude and ChatGPT. See Use bisica in Claude or ChatGPT.

Was this page helpful?