What is the RAG Engineer skill for Claude Code?

It is a specialized capability that provides Claude with domain-specific patterns for designing, implementing, and optimizing Retrieval-Augmented Generation systems.

Why is semantic chunking better than fixed-size chunking?

Semantic chunking respects document structure and sentence boundaries, preserving the meaning that arbitrary token limits often break.

What are the common anti-patterns this skill helps avoid?

It prevents common mistakes like using fixed-size chunks, indexing noisy content without filtering, and failing to measure retrieval quality independently from generation.

Does it support hybrid search implementations?

Yes, it includes patterns for combining keyword-based matching (like BM25) with semantic vector similarity using Reciprocal Rank Fusion.

How does this skill improve AI generation quality?

By focusing on 'garbage in, garbage out,' it provides strategies for high-quality retrieval, ensuring the LLM receives the most relevant and coherent context to minimize hallucinations.

RAG Engineer

Name: RAG Engineer
Author: lev-os

bylev-os

0•

Ciencia de Datos y ML

Optimizes Retrieval-Augmented Generation architectures through advanced semantic chunking, hybrid search strategies, and vector embedding pipelines.

The RAG Engineer skill transforms Claude into a specialized systems architect focused on the critical infrastructure of Retrieval-Augmented Generation. It provides expert guidance on bridging the gap between raw data and LLM generation by implementing semantic chunking, multi-level hierarchical retrieval, and hybrid search pipelines that combine vector similarity with keyword matching. This skill is essential for developers building production-grade AI agents that need to minimize hallucinations and maximize context relevance through sophisticated data preprocessing, metadata filtering, and rigorous retrieval evaluation.

Características Principales

01Hierarchical retrieval pipeline design

020 GitHub stars

03Semantic document chunking and preprocessing

04Vector embedding and similarity search implementation

05Context window and relevance optimization

06Hybrid search architecture (BM25 + Semantic)

Casos de Uso

01Optimizing AI agent memory retrieval for long-term context retention

02Building high-accuracy Q&A systems for complex technical documentation

03Improving search precision in internal enterprise knowledge bases

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lev-os/agents rag-engineer

For use in Claude.ai and ChatGPT

Download Skill