diff

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

• Third-party content exposure detected (high risk: 0.85). Yes—at runtime the required workflow navigates to the provided PROTO/BUILD URLs and runs page.evaluate(inventory, ...), which reads outsider-authored free text from the loaded pages’ DOM (e.g., headings/body/CTA labels) into the probe’s computed inventory that is then printed/used by the agent.