ai-moderating-content

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests and analyzes arbitrary user-generated content — e.g., the ModerateContent/ForumModerator and ListingModerator paths that call self.assess(...) on inputs like "content", "comment", "title"/"description" and process incoming_comments/incoming_listings — so the agent will read untrusted third-party posts/listings that could contain indirect prompt injections.