nanobanana-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 "Google検索グラウンディング" via the --search option (seen in SKILL.md, references/prompt-guide.md and in scripts/generate_image.py where payload["tools"] = [GOOGLE_SEARCH_TOOL]) which causes the model to consume live, public web search results (untrusted third‑party content) as part of its generation workflow and thus could enable indirect prompt injection.