sutui-ai

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

• Third-party content exposure detected (high risk: 0.70). The skill accepts arbitrary public image URLs (e.g., the upload_image tool and the "image_url"/"end_image_url" parameters used for 图生图/图生视频 noted in the README), meaning the agent will fetch and process untrusted third-party content provided by users or the web as part of its workflow.