← All Skills · Marketplace

RAG Pipeline Template for Agents

Complete RAG pipeline: document ingestion, chunking, embedding, retrieval, and generation — ready to run.

What It Does

End-to-end Retrieval-Augmented Generation pipeline template: document loading (PDF, HTML, Markdown), text chunking strategies (fixed-size, sentence-aware, recursive), OpenAI embedding generation, vector store upsert, semantic retrieval with context assembly, and LLM generation with source citations. Includes reranking with cross-encoders, multi-query retrieval for better recall, and evaluation patterns (precision@k, context relevance). Production-ready from day one.

What You Get

• Document loading + chunking strategy guide (fixed, sentence, recursive)
• Embedding → store → retrieve → generate pipeline code template
• Reranking, multi-query retrieval, and evaluation patterns

Install

Download rag-pipeline-template.md — no installation required

No human required after purchase. Your agent reads this page and executes the install.

Example Output

Indexed 500 documents (2,341 chunks). Query 'What is the refund policy?' → Answer with 3 source citations. Retrieval time: 45ms.

#rag#retrieval#embeddings#vector-db#llm