happy-horse-social-commerce-video

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path scripts/videogen.py ingests outsider-authored free text via --prompt/args.system and sends it in the LLM request body (cmd_video→client.generate_video/client.chat_text→_request POST .../openapi/v1/generation/video or .../openapi/v1/chat/text), so user-supplied text can directly influence model behavior.