claude-rag-skills

🥈Silver

Build, audit, and optimize RAG pipelines with Claude Code. Ideal for operations teams managing AI-driven knowledge retrieval systems. Connects to vector databases and LLM workflows for enhanced document processing and retrieval.

3000Updated 1mo ago

Intermediate30min to implementautomation

Saves ~210 min per use

Quick InstallView Source

git clone https://github.com/floflo777/claude-rag-skills.git

Works with:

Claude

Overview

About This Skill

Claude RAG Skills provides four specialized tools for RAG system development: RAG Audit scans your codebase to identify chunking issues, embedding problems, retrieval anti-patterns, and production gaps; RAG Eval measures system quality using retrieval metrics like Recall@K and NDCG plus generation metrics like faithfulness and relevance; Chunking Advisor recommends optimal strategies based on document type, query patterns, and embedding models; RAG Scaffold generates production-ready boilerplate for frameworks including LangChain, LlamaIndex, and TypeScript implementations. These skills help operations teams and developers catch issues early, benchmark performance against best practices, and accelerate RAG deployment with tested configurations.

How to Use

Install via Claude Code Marketplace with `/plugin install rag-audit`, `/plugin install rag-eval`, `/plugin install chunking-advisor`, and `/plugin install rag-scaffold`, or manually copy the skills directory to your Claude Code configuration. Invoke each skill by its command (e.g., `/rag-audit`) in any Claude Code session and answer follow-up questions about your system.

Use Cases

Identify and fix chunking anti-patterns in existing RAG implementations

Evaluate RAG system quality with standard retrieval and generation metrics

Get chunking strategy recommendations for different document types and embedding models

Generate production-ready RAG boilerplate with caching, reranking, and error handling

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/floflo777/claude-rag-skills

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

I need to build a RAG pipeline for [COMPANY] in the [INDUSTRY] sector. The pipeline should connect to [VECTOR_DATABASE] and process [DATA_TYPE] documents. Provide a step-by-step guide with Claude Code snippets to set up the pipeline, including data ingestion, vectorization, and retrieval components.

Example Output

# RAG Pipeline Setup Guide for [COMPANY]

## Step 1: Data Ingestion
```python
# Claude Code snippet for data ingestion
from langchain.document_loaders import PyPDFLoader

loader = PyPDFLoader("[DATA_PATH]/documents/")
documents = loader.load()
```

## Step 2: Vectorization
```python
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import Chroma

text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
docs = text_splitter.split_documents(documents)
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2")
vectorstore = Chroma.from_documents(docs, embeddings)
```

## Step 3: Retrieval
```python
from langchain.chains import RetrievalQA
from langchain.llms import Claude

llm = Claude(temperature=0)
qa_chain = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=vectorstore.as_retriever())
```

## Optimization Tips
- Ensure chunk size and overlap are optimized for your document type.
- Monitor and update embeddings regularly to maintain retrieval accuracy.
- Implement caching for frequently accessed documents to improve performance.