Chess MCP Server
Provides an interactive chess game experience through MCP tools with a web-based chessboard interface. Enables users to play chess games, make moves using standard algebraic notation, and manage persistent game state across sessions.
README
Chess MCP Server
A Model Context Protocol (MCP) server that provides an interactive chess game experience using MCP-UI. This server exposes chess game functionality through MCP tools and provides a web-based chessboard interface for playing games.
Features
- ♟️ Full Chess Game: Complete chess implementation using chess.js
- 🔧 MCP Tools: Chess game actions exposed as MCP tools
- 🎮 Interactive UI: Web-based chessboard using ChessBoard.js
- 💾 Game State Management: Persistent game state with session-based saves
Workshop
In the ./workshop directory, you can find a step by step to creating this MCP server.
Architecture
This MCP server is built using Express.js with integrated MCP capabilities:
Key Components
- MCP Server (
src/server.ts): Main server with chess tools and UI generation - Store (
src/lib/store.ts): Game state persistence using file system - MCP Framework (
src/lib/mcp/): Custom MCP middleware implementation with Express integration
MCP Tools Available
chess_get_board_state: Get the current state of the chess board with visual representationchess_move: Make a move in standard algebraic notation (e.g., e4, Nf3, O-O) or start a new game
Getting Started
Prerequisites
- Node.js 18+
- npm
Installation
Install the dependencies:
npm install
Development
Start the development server:
npm run dev
The server will be available at:
- MCP endpoint:
http://localhost:3000/mcp
Development Commands
npm run dev- Start development server
Usage
As an MCP Server
The server exposes MCP tools that can be consumed by MCP clients:
// Get current board state
{
"method": "tools/call",
"params": {
"name": "chess_get_board_state",
"arguments": {}
}
}
// Make a move
{
"method": "tools/call",
"params": {
"name": "chess_move",
"arguments": {
"move": "e4"
}
}
}
// Start a new game
{
"method": "tools/call",
"params": {
"name": "chess_move",
"arguments": {
"move": "new"
}
}
}
Interactive Chessboard
When using the MCP tools, each response includes:
- Text representation: ASCII board and FEN notation with valid moves
- Interactive UI: Drag-and-drop chessboard that allows moves via mouse interaction
Game Rules
- Standard chess rules apply
- Uses chess.js for move validation and game logic
- Supports all legal chess moves including castling, en passant, and pawn promotion
- Game state persists per session ID
Project Structure
src/
├── lib/
│ ├── mcp/ # Custom MCP middleware framework
│ │ ├── index.ts # Main exports
│ │ ├── middleware.ts # Express middleware integration
│ │ ├── header.ts # HTTP header utilities
│ │ └── uuid.ts # UUID validation
│ └── store.ts # Game state persistence
├── server.ts # Main server with chess tools and UI
games/ # Persistent game files (FEN notation)
Game State Persistence
- Each chess game is saved to a JSON file in the
games/directory - Games are identified by MCP session ID
- Game state is stored as FEN (Forsyth-Edwards Notation) strings
- Games automatically resume when reconnecting with the same session
Dependencies
- chess.js: Chess game logic and move validation
- @chrisoakman/chessboardjs: Interactive chessboard UI
- @mcp-ui/server: MCP UI resource creation
- @modelcontextprotocol/sdk: MCP server implementation
- express: Web server framework
License
Apache 2.0 License - see LICENSE file for details.
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。