browser

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

• Third-party content exposure detected (high risk: 1.00). The skill runs a local browser automation workflow that navigates to arbitrary URLs and executes/evaluates page JavaScript (see nav.js, eval.js, pick.js which accept user-supplied URLs/expressions and return page-derived data), so it clearly ingests untrusted public web content that could carry indirect prompt-injection payloads.