The Agent Skills Directory

[SAFE]: The skill serves as an architectural guide for RAG pipelines, providing educational content and code examples using established Python libraries.
[SAFE]: Recommendations for prompt assembly include using XML tags and clear delimiters to separate context from instructions, which is a standard security practice to mitigate indirect prompt injection.
[SAFE]: External library references (unstructured, PyMuPDF, docling, markitdown, sentence_transformers) are standard, well-known tools in the AI development ecosystem.
[SAFE]: Code snippets for Reciprocal Rank Fusion and Cross-Encoder reranking are local processing logic and do not involve dangerous command execution or network exfiltration.

rag-engineer