Barebones

This minimal, dockerized MCP server template provides an HTTP-based entry point for building your own Model Context Protocol tools. It gives you a clean starting point, a ready-to-run container, and a simple example tool you can replace with your own functionality to enable interactive tool use via MCP clients.

How to use

You connect to the MCP server from an MCP client to discover and run tools exposed by the server. In this template, you will access tools via the HTTP endpoint and then invoke them through your client’s MCP UI. Start by running the container, then connect your MCP client to the server URL to browse available tools and execute them within your chat or workflow.

Key usage patterns you can rely on includes testing the example tool, replacing it with your own tools, and validating tool schemas using an MCP inspector or your client’s tool explorer. You will see your tools listed in the client interface and will be prompted before any MCP tool is executed by the AI agent.

How to install

Prerequisites you need before installation:

• Docker is installed and running on your host
• Python 3.13 is available in the container image or environment
• You have a working network connection to pull Docker images

Step-by-step setup

# Build the MCP server image from the template
docker build -t mcp-server .
# Run the MCP server container exposing port 8000
docker run -p 8000:8000 mcp-server

Available tools

get_cat_fact

Example tool that returns a random cat fact; replace with your own tools to extend MCP capabilities