nano-banana

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly documents an "enable_google_search" input option and example ("With Google Search Grounding") which indicates the app can fetch public web/search results as real-time grounding content that the agent is expected to read and use, exposing it to untrusted third-party content that could carry indirect prompt-injection.