nmt-chat

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow ingests the outsider’s free-form chat context (the user’s messages) directly into the conversation, while grounding methodology only in the first-party canon files that are read on-demand; nonetheless, the user-provided text is an LLM input at runtime for the advisor dialogue.