gpt-image-2-seeding-image

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow only ingests outsider-authored free text when the user supplies --prompt (and optional --image files) to imagegen.py generate, which is then sent as the prompt field in client.generate_image to the API (with no intermediary selection of specific provider-authored content).