nano-banana-product-detail-page

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow ingests outsider-authored free text directly from the caller-provided --prompt argument (passed as prompt in client.generate_image(...) to /openapi/v1/generation/image), i.e., the user can submit arbitrary text for the model without any filtering/selection step.