weixin-article-reader

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

• Third-party content exposure detected (high risk: 0.85). Runtime path: the operating user supplies an outsider URL, scripts/read_weixin_article.py fetches the public HTML at that URL and extracts it into readable text (raw/js_content, then strip_tags/text) which is returned by the skill (and thus enters the agent/LLM context); this is outsider-authored web content.