raven-tools

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

• Third-party content exposure detected (high risk: 0.70). The skill calls Raven Tools via the Membrane connector to fetch reports (e.g., "Get Twitter Report", "Get Facebook Report", backlink and site-auditor reports) — these are third-party/public user-generated sources whose content the agent is expected to read and could influence subsequent action selection via Membrane actions.