# Mathematica MCP Server

<!-- mcp-name: io.github.lars20070/mathematica-mcp -->

[![codecov](https://codecov.io/gh/lars20070/mathematica-mcp/branch/master/graph/badge.svg)](https://codecov.io/gh/lars20070/mathematica-mcp)
[![Build](https://github.com/lars20070/mathematica-mcp/actions/workflows/build.yaml/badge.svg)](https://github.com/lars20070/mathematica-mcp/actions/workflows/build.yaml)
[![Python Version](https://img.shields.io/badge/dynamic/toml?url=https://raw.githubusercontent.com/lars20070/mathematica-mcp/master/pyproject.toml&query=project.requires-python&label=python)](https://www.python.org/downloads/)
[![License](https://img.shields.io/github/license/lars20070/mathematica-mcp)](https://github.com/lars20070/mathematica-mcp/blob/master/LICENSE)
[![PyPI version](https://badge.fury.io/py/mathematica-mcp.svg)](https://badge.fury.io/py/mathematica-mcp)

MCP server that wraps Mathematica's `wolframscript` command-line interface. It provides tools to evaluate Wolfram Language code and retrieve information about the Wolfram Engine installation.

| Tool | Description |
|------|-------------|
| `evaluate` | Evaluates a Wolfram Language script |
| `version_wolframscript` | Returns version of `wolframscript` |
| `version_wolframengine` | Returns version of Wolfram Engine |
| `licensetype` | Returns license type of Wolfram Engine |

## Background

* *Wolfram Language*: symbolic programming language e.g. `Integrate[x*Sin[x], x]`
* *Wolfram Engine*: kernel for running Wolfram Language code
* *WolframScript*: command-line interface to Wolfram Engine
* *Mathematica*: notebook interface to Wolfram Engine

Both Wolfram Engine and WolframScript are [freely available](https://www.wolfram.com/engine/) for personal use.

## Installation

1. Please ensure WolframScript is installed and activated on your system.

```bash
wolframscript -version
wolframscript -activate
wolframscript -code "Integrate[x*Sin[x], x]"
```

2. Install the [uv package manager](https://docs.astral.sh/uv/getting-started/installation/).

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
```

3. Edit the Claude Desktop config file and add the `mathematica-mcp` server. Note that `uvx` sets up an environment, installs the `mathematica-mcp` package, and runs the server by calling the `mathematica-mcp` entry point defined in [`pyproject.toml`](pyproject.toml). No cloning of the repository is necessary. The logs are written to the default user log directory, e.g. `~/Library/Logs/mathematica_mcp/mathematica_mcp.log` on macOS.

```json
{
  "mcpServers": {
    "mathematica-mcp": {
      "command": "uvx",
      "args": [
        "mathematica-mcp"
      ]
    }
  }
}
```
*Claude Desktop config file on macOS:* `~/Library/Application Support/Claude/claude_desktop_config.json`

4. Alternatively, you can clone the repository and then modify or extend the MCP server code. In this case, make sure to update the path to the local repository in the Claude Desktop config file.

```json
{
  "mcpServers": {
    "mathematica-mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/lars/Code/mathematica-mcp",
        "run",
        "mathematica-mcp"
      ]
    }
  }
}
```
*Claude Desktop config file on macOS:* `~/Library/Application Support/Claude/claude_desktop_config.json`