Detect hidden prompt injections before AI agents trust the web.
Scans page DOM for hidden text and invisible characters carrying instructions aimed at AI agents before the content reaches them.