droxy

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

• Third-party content exposure detected (high risk: 0.80). The skill includes actions that ingest public web content (e.g., "Create Website Resource — Create a resource from a website URL" and "Create YouTube Resource" in the Popular actions) and proxy requests via Membrane, so the agent will fetch and incorporate untrusted third-party webpages/videos as part of its workflow (SKILL.md), which could allow indirect prompt injection.