investigation-mode

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill's "Browser Verification" step instructs the agent to use agent-browser to load the user's site, take screenshots, and read the browser console/network for errors (and it also references inspecting a deployment via vercel inspect <deployment-url>), which requires fetching and interpreting open/public page content that could be untrusted and influence subsequent diagnostic actions.