RedBookSkills

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt instructs the agent to include sensitive tokens (e.g., --xsec-token XSEC_TOKEN) and similar credential-like parameters directly in generated commands, requiring the LLM to handle/output secret values verbatim and creating an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.75). 在“内容检索与互动流程”中，运行时会通过 CDP 打开小红书页面并从 window.__INITIAL_STATE__/DOM/网络响应中提取他人发布的笔记与评论文本（如 list-feeds/search-feeds/get-feed-detail/profile-snapshot/notes-from-profile/get-notification-mentions），这些属于外部作者内容进入 LLM 上下文的间接提示注入风险。