MCP Server Cookiecutter template

A Cookiecutter template for creating Model Context Protocol (MCP) servers with Python.

Table of contents:

• 1. Introduction
• 2. Quickstart
  • 2.1. Requirements
  • 2.2. Create a project
  • 2.3. Getting started with your new server
• 3. License

1. Introduction

• This template provides you a easy managable way to create your own MCP server. MCP servers can expose:
  • Tools: Functions that LLMs can execute to perform actions
  • Resources: Data sources that LLMs can access for context
  • Prompts: Reusable templates for LLM interactions
• Features:
  • Complete project structure with all necessary components pre-configured
  • Configurable transport modes (stdio, SSE)
  • Suport tools, resources and prompts.
  • Docker support for containerized deployment
  • FastMCP integration for simple decorator-based development
  • Github actions
  • Minimal dependencies

2. Quickstart

2.1. Requirements

• Python 3.12
• Cookiecutter

2.2. Create a project

# Generate a new MCP server project from the template
cookiecutter gh:ntk148v/mcp-server-template

# Or from local copy
cookiecutter path/to/mcp-server-template

# Enter your desired input

2.3. Getting started with your new server

• File structure:

{{cookiecutter.project_name}}
├── Dockerfile
├── .github
│   └── workflows
│       ├── docker.yaml
│       └── publish.yaml
├── .gitignore
├── LICENSE
├── Makefile
├── pyproject.toml
├── README.md
├── src
│   └── {{cookiecutter.module_name}}
│       ├── __init__.py
│       ├── main.py
│       └── server
│           ├── __init__.py
│           ├── prompts
│           │   ├── __init__.py
│           │   └── prompts.py
│           ├── resources
│           │   ├── __init__.py
│           │   └── resources.py
│           └── tools
│               ├── __init__.py
│               └── tools.py
└── tests
    └── .gitkeep

main.py	Main configuration
server/__init__.py	Init FastMCP server
server/tools	Tool implementations (functions that the LLM can call to perform actions)
server/resources	Resource implementations (data sources that provide context to the LLM)
server/prompts	Prompt template implementations (reusable conversation templates)
tests/	Test cases

• Edit these files according to your need: adding a new tool, a new resource, ...
• Run dev mode for testing:

$ make dev

• Install in Claude Desktop:

$ make install

3. License

Apache 2.0

---

_{Made with ❤️ by @ntk148v}