nano-banana-pro

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow ingests outsider-authored free text directly from the user-supplied --prompt argument into the AI Hive chat/text (text content) and generation/image (prompt) requests (via cmd_chat/cmd_image → AiHiveClient.chat_text/generate_image), so an outsider can POST arbitrary poison text to affect what the model receives.