generate-image

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly supports web and image grounding from Google Search (see SKILL.md "Google Search Grounding" and scripts/generate.py which adds types.GoogleSearch when --grounding/--image-grounding is used), so the agent can fetch and incorporate arbitrary public web pages and images into its generation flow, enabling indirect prompt injection from untrusted third‑party content.