Kedro RAG

Provides Kedro project data access and reasoning tools via MCP for Claude Desktop.

python

0

GitHub Stars

python

Language

5 months ago

First Indexed

2 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": {
    "sajidalamqb-kedro-mcp-rag": {
      "command": "/path/to/your/python",
      "args": [
        "/path/to/kedro-mcp-rag/kedro_mcp.py"
      ],
      "env": {
        "PYTHONPATH": "/path/to/kedro-mcp-rag/"
      }
    }
  }
}

You can run a local Kedro MCP Rag server to enable retrieval-augmented generation (RAG) capabilities with Kedro data projects. This server configuration lets Claude Desktop (or your MCP client) query a local MCP endpoint that builds a knowledge base from your Kedro setup and uses it to answer questions or assist with data science tasks.

How to use

Start by finishing the local MCP setup and then connect your MCP client to the local server. The MCP server runs as a local process that you configure and launch from your environment. When it’s running, you can ask your MCP client to perform RAG-assisted queries against your Kedro knowledge base. The MCP tools will build or refresh the knowledge DB at a temporary path if needed, and you’ll see the tools available in the client once MCP is up.

How to install

Prerequisites: Python 3.8+, Claude Desktop app.

Clone the Kedro RAG MCP project and set up the environment.
Create and activate a virtual environment.

Additional sections

Configuration and runtime steps are shown below. Follow these steps exactly to ensure the MCP server runs correctly with Claude Desktop.

plugins:
  # ... other plugins ...
  - llmstxt:
      Markdown_description: |
        Kedro is an open-source Python framework for creating reproducible, maintainable, and modular data science code. 
        # ... rest of description ...
      full_output: llms-full.txt
      sections:
        # ... sections configuration ...

MCP configuration for Kedro RAG

Configure Claude Desktop to load the local MCP server so it can use the Kedro-based knowledge base.

{
  "mcpServers": {
    "kedro-assistant": {
      "command": "/path/to/your/python",
      "args": ["/path/to/kedro-mcp-rag/kedro_mcp.py"],
      "env": {
        "PYTHONPATH": "/path/to/kedro-mcp-rag/"
      }
    }
  }
}