pongfi-research

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

• Third-party content exposure detected (medium risk: 0.30). Skill “Pongfi Research” at runtime calls TikHub MCP tools to search/traverse and then mines readable post/comment text returned by those public social platforms (e.g., discover → tools/call, then normalization includes "text"), so an outsider can submit poison by posting on the monitored platforms which the workflow will later ingest.