About the WebMCP Scanner robot

If you found WebMCPScanner in your logs: someone asked webmcp.corsen.ai to check how ready a page of your site is for AI agents. This page explains what the robot does and how to stop it.

Identification

User agent: Mozilla/5.0 (compatible; WebMCPScanner/1.0; +https://webmcp.corsen.ai/about/scanner/) (the rendering browser adds WebMCPScanner/1.0 to Chrome's own user agent). Requests come from Corsen AI's server.

What it does on your site

• One request for the page, and one each for /robots.txt, /llms.txt, /llms-full.txt and /.well-known/webmcp.
• One render of the page in a sandboxed browser, at most 250 requests, images, video and fonts skipped.
• It never submits forms, never logs in and never runs your WebMCP tools.
• Per site: at most 1 new scan per minute and 30 per hour, whoever asks.

Stop scans of your site

Write to us from the contact page with your domain: we add it to the scanner's exclusion list (subdomains included), after which any scan request for it is refused. Reports are deleted one hour after a scan at the latest.