mcp-calculator
An MCP server that integrates the MathJS library to provide AI models with advanced calculation capabilities, including support for complex numbers, matrices, and unit conversions. It supports both stdio and HTTP transports for seamless integration with clients like Claude Desktop and GitHub Copilot.
README
mcp-calculator
An MCP server that wraps MathJS to give AI models a powerful calculate tool.
Supports stdio (Claude Desktop) and HTTP Streamable transports.
Installation
npm install
npm run build
Usage
Stdio mode (default)
node dist/index.js
HTTP mode
# default port 3000
node dist/index.js --http
# custom port
PORT=8080 node dist/index.js --http
# via environment variable
TRANSPORT=http node dist/index.js
Expression examples
| Category | Expression | Result |
|---|---|---|
| Arithmetic | 2 + 3 * 4 |
14 |
| Fractions | 1/3 + 1/6 |
0.5 |
| Powers | 2^10 |
1024 |
| Trigonometry | sin(pi / 4) |
0.70710678118655 |
| Inverse trig | acos(0) |
1.5707963267949 |
| Logarithm | log(1000, 10) |
3 |
| Complex numbers | (2 + 3i) * (1 - 2i) |
8 - i |
| Statistics | mean(1, 2, 3, 4, 5) |
3 |
| Matrix determinant | det([1, 2; 3, 4]) |
-2 |
| Matrix multiply | [1,2;3,4] * [5;6] |
[[17],[39]] |
| Unit conversion | 2 km to mile |
1.2427423844747 mile |
| Constants | e ^ (i * pi) + 1 |
2.8327021... × 10^-15 + 0i ≈ 0 |
Docker
Pull and run the pre-built image from Docker Hub:
# stdio mode (pipe JSON-RPC over stdin/stdout)
docker run --rm -i bangbang93/mcp-calculator node dist/index.js
# HTTP mode on port 3000
docker run --rm -p 3000:3000 bangbang93/mcp-calculator
# custom port
docker run --rm -p 8080:8080 -e PORT=8080 bangbang93/mcp-calculator
Build locally:
docker build -t mcp-calculator .
docker run --rm -p 3000:3000 mcp-calculator
HTTP mode — usage examples
Once the server is running in HTTP mode (node dist/index.js --http), send JSON-RPC 2.0 requests with curl:
# Initialize a session
curl -X POST http://localhost:3000/ \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"curl","version":"1.0"}}}'
# Call the calculate tool
curl -X POST http://localhost:3000/ \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"calculate","arguments":{"expression":"sin(pi/4)"}}}'
Claude Desktop configuration
Add the following to your Claude Desktop claude_desktop_config.json:
{
"mcpServers": {
"calculator": {
"command": "node",
"args": ["/absolute/path/to/mcp-calculator/dist/index.js"]
}
}
}
Replace /absolute/path/to/mcp-calculator with the actual path where you cloned this repository.
Alternatively, use the Docker image so no local build is required:
{
"mcpServers": {
"calculator": {
"command": "docker",
"args": ["run", "--rm", "-i", "bangbang93/mcp-calculator", "node", "dist/index.js"]
}
}
}
GitHub Copilot CLI
Add the server to your GitHub Copilot CLI MCP configuration (~/.config/github-copilot/mcp.json on Linux/macOS, %APPDATA%\GitHub Copilot\mcp.json on Windows):
{
"mcpServers": {
"calculator": {
"command": "node",
"args": ["/absolute/path/to/mcp-calculator/dist/index.js"]
}
}
}
Or with Docker:
{
"mcpServers": {
"calculator": {
"command": "docker",
"args": ["run", "--rm", "-i", "bangbang93/mcp-calculator", "node", "dist/index.js"]
}
}
}
After saving the config, the calculate tool is available to Copilot in any chat or inline suggestion session. You can also test it directly with gh copilot suggest:
$ gh copilot suggest "calculate sin(pi/4) using the calculator tool"
opencode
Add the server to your opencode configuration (~/.config/opencode/config.json):
{
"mcp": {
"calculator": {
"type": "local",
"command": ["node", "/absolute/path/to/mcp-calculator/dist/index.js"]
}
}
}
Or HTTP mode (start the server first with node dist/index.js --http):
{
"mcp": {
"calculator": {
"type": "remote",
"url": "http://localhost:3000"
}
}
}
Development
# Run TypeScript directly (no build step)
npm run dev
# HTTP mode during development
npm run dev:http
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。