precision-math-mcp

precision-math-mcp

Provides LLMs with accurate mathematical computation through a real calculator powered by math.js, offering tools for arithmetic, algebra, calculus, unit conversion, and more.

Category
访问服务器

README

precision-math-mcp

LLMs can't do math. They predict tokens, not compute answers. Ask an LLM to multiply large numbers or solve equations, and you might get confident-sounding wrong answers.

This MCP server fixes that by giving your LLM a real calculator. It's powered by math.js, a battle-tested math library with 14k+ GitHub stars, handling everything from basic arithmetic to symbolic calculus.

All expressions use math.js syntax - see math.js documentation for full expression reference.

Features

  • 6 Mathematical Tools:

    • calculate - Evaluate mathematical expressions
    • simplify - Simplify algebraic expressions
    • derivative - Compute symbolic derivatives
    • solve - Solve equations for variables
    • format - Format numbers with precision/notation
    • convert - Convert between units
  • 3 Transport Modes:

    • stdio - Standard I/O (default, for Claude Desktop)
    • http - Streamable HTTP transport
    • sse - Server-Sent Events (backwards compatibility)

Quick Start

Run directly without installing (requires Bun, npm, yarn, or pnpm):

# Using bunx (recommended)
bunx -y @nerdo/precision-math-mcp

# Using npx
npx -y @nerdo/precision-math-mcp

# Using yarn
yarn dlx @nerdo/precision-math-mcp

# Using pnpm
pnpm dlx @nerdo/precision-math-mcp

Installation

Global Install

# Using bun
bun add -g @nerdo/precision-math-mcp

# Using npm
npm install -g @nerdo/precision-math-mcp

# Using yarn
yarn global add @nerdo/precision-math-mcp

# Using pnpm
pnpm add -g @nerdo/precision-math-mcp

Then run with:

precision-math-mcp

Local Development

git clone https://github.com/nerdo/precision-math-mcp
cd precision-math-mcp
bun install
bun run src/index.ts

Usage

Transport Modes

# Stdio mode (default)
bunx -y @nerdo/precision-math-mcp --stdio

# HTTP mode
PORT=3141 bunx -y @nerdo/precision-math-mcp --http

# HTTP mode with authentication
PORT=3141 MCP_AUTH_TOKEN=your-secret bunx -y @nerdo/precision-math-mcp --http

Environment Variables

Variable Default Description
PORT 3141 HTTP server port
MCP_AUTH_TOKEN (none) Optional auth token for HTTP
MCP_TRANSPORT stdio Transport mode: stdio, http, sse
DESCRIPTION_MODE mandatory Tool description mode: mandatory (AI must use for all math) or relaxed (optional use)

MCP Client Configuration

Claude Desktop

Add to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "precision-math": {
      "command": "bunx",
      "args": ["-y", "@nerdo/precision-math-mcp"]
    }
  }
}

Or using npx:

{
  "mcpServers": {
    "precision-math": {
      "command": "npx",
      "args": ["-y", "@nerdo/precision-math-mcp"]
    }
  }
}

Claude Code

Add to your Claude Code MCP settings:

{
  "mcpServers": {
    "precision-math": {
      "command": "bunx",
      "args": ["-y", "@nerdo/precision-math-mcp"]
    }
  }
}

Tool Examples

calculate

Evaluate mathematical expressions with optional variables.

{
  "name": "calculate",
  "arguments": {
    "expression": "sqrt(a^2 + b^2)",
    "scope": { "a": 3, "b": 4 }
  }
}
// Returns: { "result": "5", "type": "number" }

Supports:

  • Basic arithmetic: 2 + 2 * 3 → 8
  • Functions: sin(pi/4), log(100), factorial(5)
  • Complex numbers: sqrt(-1) → i
  • Matrices: [[1,2],[3,4]] * [[1,0],[0,1]]
  • Units: 2 inch to cm → 5.08 cm

simplify

Simplify algebraic expressions.

{
  "name": "simplify",
  "arguments": {
    "expression": "2x + 3x - x"
  }
}
// Returns: { "result": "4 * x" }

derivative

Compute symbolic derivatives.

{
  "name": "derivative",
  "arguments": {
    "expression": "x^3 + 2*x",
    "variable": "x"
  }
}
// Returns: { "result": "3 * x ^ 2 + 2", "simplified": "3 * x ^ 2 + 2" }

solve

Solve equations for a variable.

{
  "name": "solve",
  "arguments": {
    "equation": "x^2 - 4 = 0",
    "variable": "x"
  }
}
// Returns: { "solutions": ["2", "-2"], "type": "exact" }

format

Format numbers with specific precision or notation.

{
  "name": "format",
  "arguments": {
    "value": "pi",
    "precision": 4
  }
}
// Returns: { "result": "3.142" }

convert

Convert between units.

{
  "name": "convert",
  "arguments": {
    "value": "100 km/h",
    "toUnit": "m/s"
  }
}
// Returns: { "result": "27.77777777777778 m / s", "numericValue": 27.77777777777778 }

Testing

# Run all tests
bun test

# Run specific test file
bun test tests/tools/calculate.test.ts

Testing with MCP Inspector

npx @modelcontextprotocol/inspector bun run src/index.ts

Opens the MCP Inspector at http://localhost:6274 where you can:

  • View all available tools
  • Invoke tools with custom arguments
  • See JSON-RPC communication logs

Development

# Type check
bunx tsc --noEmit

# Run in watch mode
bun --watch run src/index.ts

License

MIT

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选