MCP Server Streamable

You can run a streamable MCP server that speaks the Model Context Protocol over HTTP or locally via a streaming transport. It exposes a simple greeting tool you can call from MCP clients to get a personalized message, making it easy to test end-to-end MCP interactions and integrate with web-based workflows.

How to use

You connect to the HTTP endpoint for remote or web-based integrations, or run the server locally to test with MCP clients on your machine. Use the greeting tool by passing a name and receive a friendly greeting in return. For local testing, you can start the server and then point your MCP client at http://127.0.0.1:8000/mcp or at the local stdio channel you configure.

How to install

Prerequisites: Python 3.8+ and the uv tool for fast package management.

Install dependencies and set up the environment step by step.

# Step 1: Navigate to the project directory
cd d:\work\projects\mcp-servers-course\mcp-server-streamable

# Step 2: Install dependencies using uv
uv sync

# Step 3: Activate the virtual environment (if not auto-activated by uv)
.venv\Scripts\activate

Additional sections

Configuration and runtime details are provided to let you run the server in your environment and connect MCP clients. You have two practical options to start the server locally and to connect remotely.

Start the server locally using uv or directly with Python. The following commands are valid ways to run the server during development.

Configuration examples

HTTP endpoint (remote or web-based usage): you can access the MCP interface at the HTTP URL and use the remote transport for MCP communications.

{
  "mcpServers": {
    "streamable-http": {
      "type": "http",
      "url": "http://127.0.0.1:8000/mcp",
      "args": []
    }
  }
}

Local standard IO (stdio) configuration for a development run: use the runtime command that starts the server in a streaming friendly way.

{
  "mcpServers": {
    "streamable-stdio": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "python", "server.py"]
    }
  }
}

Alternatively, start the server directly with Python for a straightforward local run.

{
  "mcpServers": {
    "streamable_direct_python": {
      "type": "stdio",
      "command": "python",
      "args": ["server.py"]
    }
  }
}

Usage notes and troubleshooting

If you encounter port conflicts, check the logs for the port allocation message and stop any conflicting process. Ensure the virtual environment is activated before starting the server when not using auto-activation tools. Install dependencies with the provided package manager to avoid missing modules.

Available tools

greeting

A greeting tool that takes a name parameter and returns a personalized greeting message.