wechat-article-writer

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

• Third-party content exposure detected (high risk: 1.00). The skill's Step 1 explicitly requires using a Search tool to fetch and summarize open/public third-party sources (e.g., "官方文档、X/Twitter、Reddit、技术论坛"), which the agent must read and use to compose the article, allowing untrusted user-generated content to influence actions.