openclaw-integration

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to call Genfeed MCP endpoints such as get_trending_topics, list_posts, search_articles, and get_article which return user-generated/trending content (including social platform posts and public articles) that the agent is expected to read and act on, exposing it to untrusted third-party content.