mcpField
E
easy-mcp

easy-mcp

Local
github.com/WillChangeThisLaterAI & Agents2Updated 78w ago
Repository

About easy-mcp

easy-mcp connects configured MCP servers to OpenAI Agents through a Python runner. It is for users who want to select server access and run an agent against those capabilities.

Key features

  1. 01Configure servers in servers.yaml
  2. 02Choose servers per run
  3. 03Debug server setup
  4. 04Add custom MCP servers

Use cases

  1. 01Research links with web search
  2. 02Browse local files while researching

MCP support

The agent connects to MCP servers configured in servers.yaml. Users can select which servers it accesses, and add custom servers in mcp-servers.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Project Overview

MCP servers + OpenAI agents.

Setting up the Environment

  1. Create and Activate a Virtual Environment:

    • Run the provided setup.sh script to establish a virtual environment using Python 3.13. 
      ./setup.sh
    • The script creates a virtual environment and installs all required Python packages from requirements.txt.

  2. Install Project Dependencies:

    • Ensure you have the necessary external tools and services for the MCP servers you intend to run:
      • Lynx: Required for the lynx server, which interfaces with the Lynx terminal web search tool.
      • Docker: Required for the puppeteer and github servers, enabling Chrome control and GitHub interactions, respectively.
      • Node.js: Required for the fs server, using the npx command.

Usage Guide

Running agent.py

Basic Execution

  • You can run the agent using all the servers you have configured in servers.yaml like this

    python agent.py

  • You can choose which servers the agent will have access to with:

    # gives agent access to file system (for local file browsing) and lynx (for websearch)
python agent.py --servers fs lynx

    In my experience this results in better performance since the agent is less likely to pick the wrong tool

  • Debug Mode:

    • Use the --debug flag to initialize servers without executing the agent. This is useful if you want to make sure the servers are set up correctly 
      python agent.py --debug
Configuration

The servers are configured in servers.yaml, where each server entry includes:

  • Name: Identifier for the server.
  • Description: Brief description of the server's functionality.
  • Command & Arguments: Command-line instructions to start the server.
  • Environment Variables: Any additional environment configurations needed.

You'll typically find these in github repos for MCP servers

Roll your own

You can roll your own MCP server in the mcp-servers directory

What do I use this for?

Link finding

If I am trying to figure out how to do something (like setting up QEMU) I will often use agent.py to do the initial research for me

"find links relevant to QEMU setup on ubuntu" | python agent.py --servers lynx
MCP notes...
High level

Host: Programs (Desktop, IDE, AI tool) that wants to use MCP Client: Client that maintains 1:1 connection with servers Server: Server exposing MCP capabilities

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.