rag-implementation
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of educational content and standard implementation examples for RAG. No malicious code, obfuscation, or unauthorized access patterns were found.
- [PROMPT_INJECTION]: The skill describes a RAG architecture that ingests external data, which is an inherent surface for indirect prompt injection.
- Ingestion points: Loads local documents from the
./docsdirectory using the LangChainDirectoryLoaderas shown in SKILL.md. - Boundary markers: The provided prompt templates in SKILL.md use explicit headers ('Context:', 'Question:') and negative constraints ('If you cannot answer based on the context, say "I don't have enough information."') to help ground the model.
- Capability inventory: The skill's functionality is limited to retrieval and question-answering; it does not include examples or instructions for dangerous capabilities such as shell execution, network exfiltration, or file modification.
- Sanitization: The implementation relies on standard document loading and splitting logic without additional sanitization layers for ingested content.
Audit Metadata