home / mcp / mcp servers mcp server

MCP Servers MCP Server

Reference MCP servers that provide memory and Git tooling for LLMs via MCP

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "bitflower-mcp-servers": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-memory"
      ]
    }
  }
}

You can run Model Context Protocol (MCP) servers locally or connect to them from an MCP client to give a language model access to tools and data sources in a controlled, secure way. This guide covers practical steps to start reference MCP servers and configure a client to use them.

How to use

Configure your MCP client to connect to one or more MCP servers. A typical setup includes starting a local server instance and pointing your client to the corresponding command and arguments. You can run the reference Memory server for quick testing and the Git server for repository-related operations.

How to install

Prerequisites you need before starting any MCP server:

- Node.js and npm for TypeScript-based servers

- Python and pip for Python-based servers

- A tool to run MCP servers locally (for example, uvx or npx)

Start the Memory MCP server using npx

npx -y @modelcontextprotocol/server-memory

Start the Git MCP server using uvx

uvx mcp-server-git

Alternatively, start the Git MCP server with Python if you prefer running a Python-based server

# With pip
pip install mcp-server-git
python -m mcp_server_git