MCP Calculator Server

What is MCP Calculator Server?

A Model Context Protocol (MCP) server that provides a basic calculator tool for arithmetic operations. It integrates with MCP-compatible clients and can be deployed in Smithery. The server supports HTTP, WebSocket, and stdio communication modes.

How to use MCP Calculator Server?

Install dependencies in a virtual environment, then run the server in the desired mode. For HTTP mode, set MCP_HTTP_MODE=1 and start with uvicorn server:app --host 0.0.0.0 --port 8000. For Smithery local tool integration, set MCP_STDIO_MODE=1 and LOGGING_CONFIG=stdio, then run python server.py. Use the provided shell scripts for convenience.

Key features of MCP Calculator Server

• Basic arithmetic: add, subtract, multiply, divide
• MCP‑compatible JSON‑RPC and WebSocket endpoints
• Dedicated Smithery endpoints at /mcp
• JSON schema validation and error handling
• Multiple communication modes (HTTP, WebSocket, stdio)
• Specialized entry points for deployment scenarios

Use cases of MCP Calculator Server

• Integrating a simple calculator into MCP‑enabled applications
• Running arithmetic operations in containerized environments
• Using as a local tool within Smithery workflows
• Testing and prototyping MCP server capabilities

FAQ from MCP Calculator Server

What operations does the calculator support?

Add, subtract, multiply, and divide. Division by zero is handled with an error.

How do I run the server in HTTP mode?

Set the environment variable MCP_HTTP_MODE=1 and start with uvicorn server:app --host 0.0.0.0 --port 8000, or use the provided start-container.sh script.

How do I integrate with Smithery?

For local tools, use stdio mode with MCP_STDIO_MODE=1 and LOGGING_CONFIG=stdio. For remote tools, use HTTP mode and point Smithery to the /mcp endpoint.

What environment variables are required for stdio mode?

MCP_STDIO_MODE=1 and LOGGING_CONFIG=stdio are required for Smithery local tool integration.

Can I run the server in a Docker container?

Yes, both HTTP and Smithery‑specific Dockerfiles are provided. Run with docker build and docker run using the appropriate environment variables.