douyin-subscribe

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

• Third-party content exposure detected (medium risk: 0.30). 该技能在运行时会基于用户/自动化任务中传入的抖音号去调用外部接口https://redfox.hk/story/api/dy/data/listWorkByAccount拉取作品标题/作者名等数据，并将这些字段（如content→title）直接用于 Markdown/HTML 报告与对话展示，因此存在“外部第三方作者的自由文本进入 LLM 输出”的风险链路。

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill explicitly directs the agent to create, update, list and delete recurring automation tasks (via automation_update mode=list/create/update), silently modify their embedded --accounts parameters, and immediately execute them—actions that modify the host's scheduled automation state.