florm

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly directs the agent to fetch and run Membrane actions and proxy arbitrary requests against the Florm API (e.g., "membrane action run ... list-form-responses" and "membrane request CONNECTION_ID /path/to/endpoint"), which ingests third-party/user-generated form data that the agent is expected to read and could influence subsequent actions.