generate-image

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/generate_image.py, the required runtime workflow directly ingests the user-supplied prompt argument (and an optional --input image) into message_content and sends it to OpenRouter via requests.post(...) to /api/v1/chat/completions without selecting any specific trusted item first.