rag-methodology-guide
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill describes a Retrieval-Augmented Generation (RAG) pattern. While this pattern processes external data which creates an indirect prompt injection surface, the provided template includes boundary markers and explicit instructions to restrict the model's output to the provided context, following industry standard safety practices for RAG systems.
- [EXTERNAL_DOWNLOADS]: The skill references standard, well-maintained libraries including langchain, chromadb, and sentence-transformers for local text processing and vector storage.
Audit Metadata