gemini-3-pro-image

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

• Third-party content exposure detected (low risk: 0.10). The workflow downloads remote reference images via HTTP(S) URLs (--image-url), which can expose the agent to untrusted external content if an attacker-controlled URL is supplied.

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

• Potentially malicious external URL detected (high risk: 0.70). The script fetches a user-supplied URL via --image-url at runtime using an arbitrary, user-selected external source, representing a weak runtime external dependency.