gpt-image-2

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

• Third-party content exposure detected (medium risk: 0.30). The workflow ingests outsider-authored free text at runtime via the prompt field in the user-provided JSON passed to runcomfy run openai/gpt-image-2/text-to-image / .../edit, which is then POSTed to RunComfy’s Model API.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's edit endpoint requires user-supplied HTTPS image URLs that the RunComfy model server fetches at runtime (e.g., "https://...subject.jpg", "https://...room.jpg"), and those fetched images directly influence model outputs (image-based prompt injection risk).