wechat-title

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

• Third-party content exposure detected (high risk: 0.75). scripts/fetch_official_account_trends.py 在运行时调用外部API获取的结果字段（如文章标题/作者等）会生成Markdown文本（generate_markdown_output），然后被后续“分析爆款标题并生成新标题”的LLM步骤读取；该API返回内容属于非操作用户、非官方组织编写的外部免费文本，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: 0.70). The skill's runtime script calls the external API https://redfox.hk/story/api/cozeSkill/getWxCozeSkillData to fetch data that the agent injects into its analysis/prompting pipeline (scripts/fetch_official_account_trends.py uses this URL at runtime), so the external content directly controls the agent's outputs.