smooch

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to run Membrane actions and proxy requests against the Smooch (Sunshine Conversations) API (e.g., "membrane action run" and "membrane request CONNECTION_ID /path/to/endpoint"), which will fetch user-generated conversation messages and other external third‑party data that the agent is expected to read and act on, enabling indirect prompt injection risk.