vector-database

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests outsider-authored free text when running RAG ingestion (chunk → embed → store → retrieve), and the checklist explicitly covers “Any agent/worker that takes text → produces a vector → writes to a vector store” rather than restricting to first-party/trusted inputs.