image-2-character-consistency

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

• Third-party content exposure detected (medium risk: 0.30). Outsider-authored free text from the user-supplied --prompt (and optional --system) is sent into the LLM at runtime via cmd_image() → AiHiveClient.generate_image() which posts {"prompt": prompt} to /openapi/v1/generation/image (and cmd_chat() would do the same for chat), so the workflow ingests arbitrary user text directly.