websocket-engineer

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow is a WebSocket/Socket.IO server that ingests outsider-authored free text from connected clients via the socket.on("message", ({ roomId, text }) => ...) handler and then emits that text back into the application context (and potentially into any LLM downstream), so there is a direct indirect prompt-injection path from client-supplied text.