nano-banana-pro

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly shows enabling "google_search" grounding (see the "With Google Search Grounding" example and the Multi-Turn Conversations examples that set config tools=[{"google_search": {}}]), which causes the agent to fetch and interpret public web content that can influence generation behavior and thus exposes it to untrusted third‑party content.