The Agent Skills Directory

[SAFE]: The skill provides structured guidance and Python code for auditing RAG retrieval systems. It includes instructions for identifying failure stages like ingestion, parsing, and chunking while explicitly requiring adherence to Access Control Lists (ACLs) to prevent unauthorized data access. The provided Python snippets for calculating Recall@K and MRR metrics use standard libraries and do not contain malicious code, unauthorized network requests, or dangerous command execution. All external references link to well-known technical documentation and community resources for RAG frameworks such as LangChain, LlamaIndex, and Ragas.

rag-retrieval-audit