What chunking strategies does this skill support?

It covers a wide range of strategies including fixed-size, sentence-based, paragraph-based, recursive, and semantic chunking to ensure optimal data structure for retrieval.

Is this suitable for production environments?

Absolutely, it includes production-ready patterns for caching, streaming retrieval, fallback mechanisms, and cost optimization for embedding APIs.

Does it support query optimization?

Yes, it includes advanced techniques like Hypothetical Document Embeddings (HyDE), multi-query generation, and step-back prompting to improve retrieval recall.

Can I use this with specific vector databases?

Yes, it provides detailed guidance and implementation patterns for major databases like Pinecone, Weaviate, Qdrant, Chroma, and pgvector.

How does this skill help with RAG evaluation?

It integrates frameworks like RAGAS and provides metrics for faithfulness, context relevance, and answer correctness to systematically validate your system.

RAG Architect

Name: RAG Architect
Author: alirezarezvani

byalirezarezvani

•

5,225

•

データサイエンスとML

Designs and implements production-grade Retrieval-Augmented Generation pipelines with advanced chunking, retrieval, and evaluation strategies.

The RAG Architect skill empowers developers to build sophisticated, scalable, and accurate AI-powered retrieval systems within Claude Code. It provides deep technical guidance on document processing, embedding model selection, vector database integration (including Pinecone, Weaviate, and pgvector), and hybrid retrieval methods. Whether you are optimizing for latency, cost, or precision, this skill covers advanced implementation patterns like HyDE query transformation, reciprocal rank fusion, and automated evaluation frameworks to ensure your RAG application is enterprise-ready.

主な機能

015,225 GitHub stars

02Hybrid retrieval implementation combining dense semantic and sparse keyword search

03Comprehensive evaluation frameworks for faithfulness and context relevance

04Query transformation techniques like HyDE and Multi-query generation

05Vector database optimization for Pinecone, Weaviate, Qdrant, and pgvector

06Advanced document chunking strategies including semantic and recursive splitting

ユースケース

01Optimizing existing RAG pipelines to reduce hallucinations and improve answer accuracy

02Building high-precision internal knowledge bases for complex technical documentation

03Designing cost-effective, scalable vector search architectures for production apps

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add alirezarezvani/claude-skills rag-architect

For use in Claude.ai and ChatGPT

Download Skill