muapi-storyboard-to-cooking-video

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly fetches a user-supplied image via the person_image image_url (Step 1 & Step 2) and the agent must read and use that untrusted third-party image as an identity/reference anchor that materially directs the reference-sheet and final video generation, creating a pathway for injected visual/textual content in the image to influence behavior.