mcphost

MCP.Pizza Chef: mark3labs

This is a command-line program you install on your computer to chat with an AI model — Claude, ChatGPT, Gemini, or a local Ollama model — while it uses outside tools to look things up, read files, or run commands for you. It connects to any Model Context Protocol tool source, so its abilities depend on what you connect. Setup means installing a free tool called Go and typing commands in a terminal; cloud models also need an account and a personal access key. The creators have stopped updating this project and now recommend a newer tool called Kit.

Archived · The author made this repository read-only on GitHub.
Coding
Data
Files/PDF

Use This MCP client To

Chat with Claude, GPT, Gemini, or Ollama from one command-line app Let my AI read, search, and edit files on my computer Have my AI run terminal commands and report back the results Give my AI access to a running to-do list Connect my AI to web tools to look things up live Switch between different AI models without changing how I work

README

⚠️ MCPHost is No Longer Actively Maintained

Active development on MCPHost has been stopped. This project has been succeeded by Kit, which builds on the foundations of MCPHost with a more powerful and extensible architecture.

👉 We recommend all users migrate to Kit.

This repository is now archived and will not receive further updates or bug fixes.


MCPHost 🤖

A CLI host application that enables Large Language Models (LLMs) to interact with external tools through the Model Context Protocol (MCP). Currently supports Claude, OpenAI, Google Gemini, and Ollama models.

Discuss the Project on Discord

Table of Contents

Overview 🌟

MCPHost acts as a host in the MCP client-server architecture, where:

  • Hosts (like MCPHost) are LLM applications that manage connections and interactions
  • Clients maintain 1:1 connections with MCP servers
  • Servers provide context, tools, and capabilities to the LLMs

This architecture allows language models to:

  • Access external tools and data sources 🛠️
  • Maintain consistent context across interactions 🔄
  • Execute commands and retrieve information safely 🔒

Currently supports:

  • Anthropic Claude models (Claude 3.5 Sonnet, Claude 3.5 Haiku, etc.)
  • OpenAI models (GPT-4, GPT-4 Turbo, GPT-3.5, etc.)
  • Google Gemini models (Gemini 2.0 Flash, Gemini 1.5 Pro, etc.)
  • Any Ollama-compatible model with function calling support
  • Any OpenAI-compatible API endpoint

mcphost FAQ

Which apps does this work in?
This is its own stand-alone command-line app, not an add-on for Claude Desktop, Cursor, or ChatGPT — you open a terminal window and chat with it directly, and it can connect to Claude, OpenAI's models, Google Gemini, or a local Ollama model.
Do I need an account or a key to use it?
Yes, for the most part. For Claude, OpenAI, or Gemini you need an account and a personal access key from that company; if you use a local Ollama model instead, no account or key is required.
How hard is this to set up?
It takes some technical comfort — you'll install a free programming tool called Go, run a command to install the app, and edit a small settings file, so it's best suited to someone comfortable using a terminal.
Can I use this to have my AI read and edit files on my computer?
Yes — with the right tool connected, you can ask it to look through, read, or change files in a folder you point it to.
Can I use this to run commands or repeat a series of steps automatically?
Yes — it can run terminal commands and follow a saved script of steps, which is handy for repeating the same task without retyping it each time.
Is this project still being updated?
No — the creators have stopped maintaining it and now recommend switching to their newer tool, called Kit, for ongoing support and fixes.
Do I need to know how to code to use it?
You don't need to write code to chat with it, but getting it installed and configured does require some comfort typing commands in a terminal.