xiaohongshu-note-analyzer

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

• Third-party content exposure detected (low risk: 0.10). 运行时路径是 skill 通过用户提供的 --keyword 组装请求体并对 https://doubaoya.com/api/apis/xiaohongshu/xiaohongshu-coze/call 发起 POST，仅读取 doubaoya.com 返回的 data.items[].item（笔记标题、点赞数等）而不是直接读取任意用户粘贴的文本/邮件内容/工单正文。

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill calls the runtime endpoint https://doubaoya.com/api/apis/xiaohongshu/xiaohongshu-coze/call to fetch JSON and SKILL.md explicitly instructs the agent to relay any returned "notice" verbatim, meaning remote content can directly influence agent behavior.