Cal Server

What is Cal Server?

Cal Server is a lightweight mathematical expression calculation service built with FastMCP and the Bun runtime. It uses the expr-eval library to parse and evaluate user-input expressions and communicates via standard input/output (stdio), serving as a simple, efficient calculation tool with built-in constants.

How to use Cal Server?

Install automatically via Smithery (npx -y @smithery/cli install @pwh-pwh/cal-mcp --client claude) or manually (clone the repository, run bun install). Configure your MCP client with the command bunx cal-mcp. Start the service with bun run index.ts and invoke the cal tool with a string expression parameter (exp) to get the result.

Key features of Cal Server

• Supports mathematical expression calculation via the cal tool
• Includes built-in constants: E, PI, true, false
• Accepts a single string expression parameter
• Built on FastMCP for MCP protocol compatibility
• Runs on the Bun runtime environment
• Communicates exclusively over stdio

Use cases of Cal Server

• Evaluating arithmetic expressions within an AI agent conversation
• Computing results that use constants like PI or E without manual lookup
• Performing quick math operations in a lightweight, serverless context

FAQ from Cal Server

What runtime is required for Cal Server?

Cal Server requires the Bun runtime (latest version recommended). You can install it from https://bun.sh/.

How does Cal Server communicate?

It communicates over standard input and output (stdio). It does not support HTTP or other transports.

What built-in constants does Cal Server support?

The server includes the constants E, PI, true, and false, which map to Bun’s Math.E, Math.PI, true, and false.

What happens if an invalid expression is entered?

If the input is not a valid mathematical expression, Cal Server may throw an error. Ensure expressions are syntactically correct.

What are the known limits of Cal Server?

The service only supports stdio transport and does not include authentication. Expressions must be mathematically valid; otherwise an error may occur.