About the SiteCheck scanner

SiteCheck fetches one public page when a person asks for it, on this website or through the SiteCheck plugin in ChatGPT. It is not a crawler: it never follows links to other pages, and it only visits a page someone typed in.

What a scan does

One HTTP request for the page (User-Agent SiteCheckBot/1.0 (+https://sitecheck.datacleave.com/about-bot)), plus small requests for /robots.txt, /sitemap.xml, the favicon and the share image named in the page's meta tags.
One load of the page in a headless Chromium browser to run the axe-core accessibility rules. The browser identifies itself with the same SiteCheckBot/1.0 token in its User-Agent.
One request to Google's PageSpeed Insights service, which runs its own Lighthouse test of the page from Google's infrastructure.

Each page is scanned at most once per 24 hours unless someone asks for a rescan, and a rescan of the same page is allowed every 10 minutes at most.

How to opt out

Any of these stops future scans of your site. The first two are checked before every scan, so they take effect immediately.

Add to your robots.txt:
```
User-agent: SiteCheckBot
Disallow: /
```
Add to the page's <head>: <meta name="sitecheck" content="noscan">
E-mail support@datacleave.com from an address at your domain; we add the host to our opt-out list and remove any existing report page.

Report pages

A scan result has a page at https://sitecheck.datacleave.com/r/<host>. These pages are noindex by default. A host's page becomes indexable only after it has been scanned three or more times, and never for a host that has opted out or whose page is itself marked noindex. Report pages show the scan's findings and the page's own public meta tags and share image.