omni-image

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the agent to perform web/image searches and download reference photos from public websites (see "Finding reference photos" and "Agent behavior" sections), meaning the agent fetches and ingests untrusted third‑party content that it will read and use to drive image-editing actions.