Which vector databases does this RAG skill support?

The skill provides implementation patterns for leading vector stores including Pinecone, Weaviate, Milvus, Chroma, Qdrant, and FAISS.

How does this skill help reduce AI hallucinations?

By implementing grounding prompts and source-backed retrieval, the skill ensures the LLM relies on provided context rather than its training data alone.

Does it include strategies for handling long documents?

Absolutely. It covers various chunking strategies, including recursive splitting and semantic chunking, to ensure data fits within LLM context windows effectively.

Can I implement hybrid search with this skill?

Yes, it includes specific patterns for combining dense semantic search (embeddings) with sparse keyword search (BM25) for superior retrieval accuracy.

RAG Implementation for LLM Apps

Name: RAG Implementation for LLM Apps
Author: HermeticOrmus

byHermeticOrmus

数据科学与机器学习

Builds high-performance Retrieval-Augmented Generation systems using vector databases and semantic search to ground AI responses in external knowledge.

关于

This skill provides a comprehensive framework for building production-grade Retrieval-Augmented Generation (RAG) systems. It guides developers through the entire pipeline—from sophisticated document chunking and embedding generation to vector database configuration and advanced retrieval strategies like hybrid search and reranking. By implementing these patterns, you can build LLM applications that provide accurate, factual responses based on proprietary documentation or external knowledge bases while significantly reducing hallucinations and improving source transparency.

主要功能

0 GitHub stars
Comprehensive vector database integration (Pinecone, Weaviate, Chroma, Qdrant, and FAISS)
Advanced retrieval patterns including Hybrid Search, Multi-Query, and Contextual Compression
Sophisticated chunking strategies for recursive, token-based, and semantic document splitting
Standardized evaluation metrics for measuring accuracy, retrieval quality, and groundedness
Reranking methods using Cross-Encoders and Maximal Marginal Relevance (MMR) for result quality

使用场景

Creating real-time chatbots that access current, factual information from external databases
Building Q&A systems over proprietary internal company documents and wikis
Developing research tools that provide automated citations and evidence-based answers

关于

主要功能

0 GitHub stars
Comprehensive vector database integration (Pinecone, Weaviate, Chroma, Qdrant, and FAISS)
Advanced retrieval patterns including Hybrid Search, Multi-Query, and Contextual Compression
Sophisticated chunking strategies for recursive, token-based, and semantic document splitting
Standardized evaluation metrics for measuring accuracy, retrieval quality, and groundedness
Reranking methods using Cross-Encoders and Maximal Marginal Relevance (MMR) for result quality

使用场景

Creating real-time chatbots that access current, factual information from external databases
Building Q&A systems over proprietary internal company documents and wikis
Developing research tools that provide automated citations and evidence-based answers