nanobanana-2-image-generation

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly supports grounding with Google Search / Google Image Search (see SKILL.md "Enable grounding via --google-search", references/image-generation-api.md section "Grounding (Web Search + Image Search)"), and the CLI script scripts/nanobanana-cli.js implements a --google-search flag that adds a googleSearch tool to the model payload, which causes the agent to ingest and act on public web/image search results—exposing it to untrusted third-party content that could carry indirect prompt-injection.