topic-participation

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

• Third-party content exposure detected (high risk: 0.80). The skill's Step 1 "Trend Discovery and Monitoring" explicitly instructs checking and using open/public, user-generated sources (Xiaohongshu Hot Topics/Explore pages, comment sections, community groups, Weibo, Douyin, Google Trends, etc.), which the agent would read and interpret to make posting and engagement decisions—exposing it to untrusted third‑party content that could indirectly inject instructions.