Claude RAG

Provides a RAG pipeline with MCP integration to access local documents via Claude Desktop for seamless retrieval and chat.

python

GitHub Stars

python

Language

6 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "kenjisekino-claude-rag-mcp-pipeline": {
      "command": "/path/to/your/project/rag_env/bin/python",
      "args": [
        "/path/to/your/project/mcp_server.py"
      ],
      "env": {
        "OPENAI_API_KEY": "YOUR_OPENAI_API_KEY",
        "ANTHROPIC_API_KEY": "YOUR_ANTHROPIC_API_KEY"
      }
    }
  }
}

You can run a Retrieval Augmented Generation (RAG) pipeline that integrates Model Context Protocol (MCP) to access your local documents from Claude Desktop. This setup lets you search, summarize, and chat about your documents directly through Claude while keeping data local and under your control.

How to use

Use the MCP-enabled RAG server with Claude Desktop to access your own documents during chats. You can ingest documents, perform semantic search, and ask questions that combine your documents with general AI capabilities. When you prompt Claude, it can automatically search your documents for relevant details and cite sources from your local collection. To start a chat, simply open Claude Desktop, ensure your MCP server is running, and begin a conversation about your documents.

How to install

Prerequisites: Python 3.8 or newer, OpenAI API key, Anthropic API key, and Claude Desktop installed for MCP integration.

# 1) Clone the project repository
git clone https://github.com/kenjisekino/claude-rag-mcp-pipeline.git
cd claude-rag-mcp-pipeline

# 2) Create a Python virtual environment
python3 -m venv rag_env
source rag_env/bin/activate

# 3) Install dependencies
pip install -r requirements.txt

# 4) Configure environment variables
cp .env.example .env
# Edit .env and insert your API keys (OPENAI_API_KEY, ANTHROPIC_API_KEY)

# 5) Create the documents folder (for your content)
mkdir documents

# 6) Start the Streamlit web interface for document ingestion and chat
streamlit run app.py

Additional sections

MCP Configuration for Claude Desktop is provided below. This enables native document access through the Claude Desktop environment and wires your local project MCP server into Claude’s tooling.

{
  "mcpServers": {
    "personal-documents": {
      "command": "/path/to/your/project/rag_env/bin/python",
      "args": ["/path/to/your/project/mcp_server.py"],
      "env": {
        "OPENAI_API_KEY": "your_key",
        "ANTHROPIC_API_KEY": "your_key"
      }
    }
  }
}

Start the MCP server

python mcp_server.py

Use Claude Desktop — ensure personal-documents is enabled under Search and tools and begin a chat that prompts document search when relevant.

Built by

VeilStrat

AI signals for GTM teams