mcp-local-client-server

sjha3/mcp-local-client-server

3.1

If you are the rightful owner of mcp-local-client-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Model Context Protocol (MCP) server and client project facilitates interaction between AI agents and local tools or data sources.

MCP Client-Server Project

This project provides a local implementation of an MCP (Model Context Protocol) server and client, enabling AI agents to interact with local tools and data sources.

Features

1. MCP Server

The MCP server exposes tools to perform the following operations:

  • Operations in local files: Read and process data from local CSV and Parquet files.
  • Get financial data: Retrieve financial data using the Yahoo Finance API.
  • Manage lifecycle of an object in Kusto database: Create, update, and manage objects in an Azure Data Explorer (Kusto) database.

2. MCP Client

The MCP client provides the following capabilities:

  • Connects to the MCP server created above.
  • Creates AI agents using Autogen: Utilizes the Autogen framework to instantiate multiple AI agents.
  • Assigns tools from the MCP server to these agents: Each agent is equipped with the appropriate tools for its tasks.

How to Run

  1. Install Python dependencies using pip: 
    pip install -r requirements.txt
  2. Create .env file and add these environment variables: 
    azure_endpoint
azure_deployment
api_version
api_key
azure_model_name
  3. Run the MCP server: 
    python server-main.py
  4. Run the MCP client: 
    python client-main.py

This will bring up a prompt where you can chat with the AI agents.

Based on your question, the appropriate agent will be invoked and the requested operation will be performed automatically.