nano-banana-pro-product-retouch

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

• Third-party content exposure detected (medium risk: 0.30). The runtime ingests outsider-authored free text provided at invocation time via the --prompt CLI argument, and forwards it to the AI Hive OpenAPI /openapi/v1/chat/text or /openapi/v1/generation/image endpoints without any prior item selection (e.g., skill_generate → cmd_image → client.generate_image(..., prompt=...)).