ai-copy-editing

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

• Third-party content exposure detected (medium risk: 0.65). 该技能在运行时会接收用户提供的 materialUrl（公开可访问文章/页面/竞品链接）并将其内容作为参考材料读入 LLM 上下文；该 URL 对应的网页属于“公共 web 内容”这一类外部来源，存在间接提示注入风险。

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

• Potentially malicious external URL detected (high risk: 1.00). The provided runner makes runtime calls to the external execute API at https://ai-skills.ai (POST to /api/execute) to run the skill remotely, meaning the skill relies on that external service which executes code/prompts during runtime.