- Home
- MCP servers
- Simple
Simple
- python
1
GitHub Stars
python
Language
5 months ago
First Indexed
2 months ago
Catalog Refreshed
Documentation & install
Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.
You can run a lightweight MCP server that exposes math tools over a streamable HTTP transport. It follows the MCP protocol and provides a simple, well-documented starting point for learning and building remote MCP services that can be deployed to cloud platforms.
How to use
You connect to the MCP server with a client that speaks the MCP protocol. After the server initializes, you can request the available tools, call a tool with arguments, and receive the computed results in a streaming, structured format. Use a session-based flow: initialize, acknowledge initialization, then perform tool lookups and calls. The server exposes two math tools, add and multiply, which you can invoke from your client.
How to install
Prerequisites: Python and a modern HTTP client or MCP client library. You will also need the ability to run commands in your shell and access to a network to connect to the MCP endpoint.
# 1) Clone the example MCP server repository
git clone https://github.com/oleksandrsirenko/mcp-simple-server.git
auth
# 2) Change into the project directory
cd mcp-simple-server
# 3) Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate
# 4) Install dependencies
pip install -r requirements.txt
# 5) Run the MCP server
python main.py
# The server starts and listens for MCP requests at port 8000 by default
Additional content
Configuration and deployment are described for several environments. You can run the server locally, or deploy to Railway, Heroku, or Render. For remote operation, you can use a CLI helper to connect to a deployed MCP endpoint, or run your own proxy for debugging. The server provides a local URL at http://localhost:8000/mcp/ when running locally.
Deployment notes
Local testing uses the URL http://localhost:8000/mcp/. Remote deployments are commonly tested with a script that exercises the MCP protocol against your deployed endpoint. When deploying, you can use common cloud platforms that auto-detect deployment configurations.
{
"mcpServers": {
"simple-http": {
"type": "http",
"name": "simple_http",
"url": "http://localhost:8000/mcp/",
"args": []
}
}
}
Claude Desktop integration guidance
You can configure Claude Desktop to connect to your MCP server so Claude can call tools like add and multiply. If you run the server locally, you can use a local configuration that points to your local MCP URL. For remote servers, you can use the remote integration flow provided by the Claude Desktop plugin to point to your deployed MCP URL with the appropriate headers for streaming responses.
Troubleshooting and notes
If the server won’t start, verify that port 8000 (or the port you configured) is available, and check for syntax or dependency errors in the startup output. If protocol errors occur, run automated tests that exercise initialize, tool listing, and tool calls to surface mismatches between client expectations and server behavior. Ensure the correct MCP endpoint path ending with a trailing slash is used, typically /mcp/.
Available tools
add
Adds two numbers together and returns the result.
multiply
Multiplies two numbers and returns the result.