Does it support hybrid search?

Yes, it includes advanced patterns for combining semantic vector search with keyword-based (BM25) search to capture both conceptual meaning and specific terminology.

What is RAG and why is it used?

Retrieval-Augmented Generation (RAG) is a technique that provides an LLM with specific, external data to ensure its answers are accurate, up-to-date, and grounded in your own documentation rather than just its training data.

How does this skill help with AI hallucinations?

By implementing strict retrieval patterns and grounding prompts, this skill ensures the AI only answers based on the provided context and can include citations to source material.

Can I use this for production applications?

Absolutely. The skill includes best practices for reranking, metadata filtering, and evaluation metrics specifically designed for production-grade reliability.

Which vector databases can I use with this skill?

This skill provides implementation patterns for several popular options including Pinecone (managed), Weaviate (hybrid), Chroma (local/lightweight), and FAISS (local library).

RAG Implementation

Name: RAG Implementation
Author: Microck

byMicrock

•

データサイエンスとML

Implements Retrieval-Augmented Generation (RAG) systems using vector databases and semantic search to ground LLM applications in external knowledge.

概要

This skill provides a comprehensive framework for building production-ready RAG systems, enabling Claude to architect applications that access and reason over proprietary or domain-specific datasets. It guides users through the entire RAG lifecycle, including document ingestion, sophisticated chunking strategies, embedding generation, and advanced retrieval patterns such as hybrid search and semantic reranking. By leveraging these patterns, developers can significantly reduce hallucinations and build high-accuracy Q&A systems, documentation assistants, and research tools that provide verifiable source citations.

主な機能

Advanced document chunking and recursive splitting strategies
Comprehensive evaluation metrics for groundedness and retrieval quality
Hybrid search patterns combining semantic and keyword retrieval
Multi-vector store integration including Pinecone, Chroma, and Weaviate
81 GitHub stars
Contextual compression and reranking for high-precision results

ユースケース

Reducing AI hallucinations by grounding responses in verified knowledge bases
Building enterprise-grade Q&A systems over private document repositories
Creating documentation assistants with automated source citations

概要

主な機能

Advanced document chunking and recursive splitting strategies
Comprehensive evaluation metrics for groundedness and retrieval quality
Hybrid search patterns combining semantic and keyword retrieval
Multi-vector store integration including Pinecone, Chroma, and Weaviate
81 GitHub stars
Contextual compression and reranking for high-precision results

ユースケース

Reducing AI hallucinations by grounding responses in verified knowledge bases
Building enterprise-grade Q&A systems over private document repositories
Creating documentation assistants with automated source citations