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 ./docs directory using the LangChain DirectoryLoader as 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
Risk Level
SAFE
Analyzed
Mar 28, 2026, 08:17 PM
Security Audit — agent-trust-hub — rag-implementation