phoenix-channels-essentials

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow is Phoenix channel handling (connect/3, join/3, handle_in/3) and the only ingested free text shown is client-sent payload fields (e.g., handle_in("new_message", %{"body" => body}, ...)), which are outsider-authored messages from other users/clients entering the LLM context via the agent’s own message-handling path.