menu-demand-radar

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

• Third-party content exposure detected (high risk: 0.85). At runtime the skill calls UnifAPI endpoints that fetch and ingest public web/AI/social content (e.g., geo/serp captures AI answer text + cited sources; tiktok/search / tiktok/hashtags/{id}/videos read public video metadata/captions), which is outsider-authored free text that can be fed into the agent’s LLM context.