zerodha-claude-mcp
Enables AI models to interact with the Zerodha trading platform for portfolio management, trading, and account operations through natural language.
README
🚀 Zerodha Trading MCP Server -- Claude
<div align="center">
A powerful Model Context Protocol (MCP) server that enables AI models to interact seamlessly with the Zerodha trading platform
</div>
📋 Table of Contents
- ✨ Features
- 🎯 What is MCP?
- ⚡ Quick Start
- 📦 Installation
- 🐳 Docker Installation
- ⚙️ Configuration
- 🔧 Claude Desktop Setup
- 📖 API Reference
- 🛠️ Development
- 🤝 Contributing
- 📄 License
- ⚠️ Disclaimer
✨ Features
| Feature | Description |
|---|---|
| 🔐 Secure Authentication | Robust authentication with Zerodha API using official KiteConnect SDK |
| 📊 Portfolio Management | Real-time access to your complete portfolio holdings and valuations |
| 📈 Position Tracking | Monitor open positions with live P&L calculations |
| 💹 Order Execution | Place buy/sell orders with market/limit options |
| 👤 Profile Access | Retrieve comprehensive user profile and account information |
| 🔄 Real-time Data | Live market data and portfolio updates |
| 🛡️ Type Safety | Built with TypeScript and Zod validation for reliability |
🎯 What is MCP?
The Model Context Protocol (MCP) is a standardized way for AI models to interact with external systems and APIs. This server acts as a bridge between AI assistants (like Claude) and the Zerodha trading platform, enabling natural language trading operations.
⚡ Quick Start
Prerequisites
Before you begin, ensure you have:
- ✅ Node.js (v16 or higher) - Download here
- ✅ Zerodha Trading Account - Sign up here
- ✅ Zerodha API Access - Get API credentials
- ✅ Bun Runtime (recommended) - Install here
📦 Installation
Step 1: Clone the Repository
# Clone the repository
git clone https://github.com/gamandeepsingh/zerodha-claude-mcp.git
# Navigate to the project directory
cd zerodha-claude-mcp
Step 2: Install Dependencies
Choose your preferred package manager:
# Using npm
npm install
# Using bun (recommended for better performance)
bun install
# Using yarn
yarn install
Step 3: Environment Setup
Create a .env file in the root directory:
# Copy the example environment file
cp .env.example .env
Add your Zerodha credentials to the .env file:
# Zerodha API Configuration
KITE_API_KEY=your_api_key_here
KITE_SECRET_KEY=your_secret_key_here
REQUEST_TOKEN=your_request_token_here
ACCESS_TOKEN=your_access_token_here
# Optional: Debug mode
DEBUG=false
💡 Pro Tip: You can get your API credentials from the Kite Connect Developer Console
🐳 Docker Installation (Alternative)
For a containerized setup, you can use Docker to run the MCP server:
Prerequisites for Docker
- ✅ Docker - Install Docker
- ✅ Docker Compose (optional, for easier management) - Usually included with Docker Desktop
Option 1: Using Docker Compose (Recommended)
-
Clone the repository:
git clone https://github.com/gamandeepsingh/zerodha-claude-mcp.git cd zerodha-claude-mcp -
Create environment file:
# Create .env file with your credentials cp .env.example .env -
Edit the .env file with your Zerodha credentials:
KITE_API_KEY=your_api_key_here KITE_SECRET_KEY=your_secret_key_here REQUEST_TOKEN=your_request_token_here ACCESS_TOKEN=your_access_token_here DEBUG=false -
Build and run with Docker Compose:
# Build and start the container docker-compose up --build # Run in detached mode (background) docker-compose up -d --build -
Stop the container:
docker-compose down
Option 2: Using Docker directly
-
Build the Docker image:
docker build -t zerodha-claude-mcp . -
Run the container:
docker run -d \ --name zerodha-mcp \ -e KITE_API_KEY=your_api_key \ -e KITE_SECRET_KEY=your_secret_key \ -e REQUEST_TOKEN=your_request_token \ -e ACCESS_TOKEN=your_access_token \ zerodha-claude-mcp
Docker Management Commands
# View running containers
docker ps
# View logs
docker logs zerodha-mcp
# Stop the container
docker stop zerodha-mcp
# Remove the container
docker rm zerodha-mcp
# Remove the image
docker rmi zerodha-claude-mcp
Claude Desktop Configuration with Docker
When using Docker, update your claude_desktop_config.json to point to the containerized version:
{
"mcpServers": {
"zerodha-trading": {
"command": "docker",
"args": ["exec", "-i", "zerodha-mcp", "bun", "run", "index.ts"],
"env": {
"KITE_API_KEY": "your_actual_api_key",
"KITE_SECRET_KEY": "your_actual_secret_key",
"REQUEST_TOKEN": "your_actual_request_token",
"ACCESS_TOKEN": "your_actual_access_token"
}
}
}
}
⚠️ Note: Make sure your Docker container is running before starting Claude Desktop when using this configuration.
⚙️ Configuration
Getting Zerodha API Credentials
- Sign up for a Kite Connect app at developers.kite.trade
- Create a new app and note down your
api_keyandapi_secret - Generate tokens using the authentication flow
- Add credentials to your
.envfile
Token Generation Process
# Step 1: Get the login URL
curl "https://kite.trade/connect/login?api_key=your_api_key&v=3"
# Step 2: After login, extract request_token from redirect URL
# Step 3: Generate access_token using the request_token
🔧 Claude Desktop Setup
To integrate this MCP server with Claude Desktop, follow these steps:
Step 1: Locate Claude Desktop Config
Find your claude_desktop_config.json file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
Step 2: Add MCP Server Configuration
Add the following to your claude_desktop_config.json:
{
"mcpServers": {
"zerodha-trading": {
"command": "bun",
"args": ["run", "/absolute/path/to/your/index.ts"],
"env": {
"KITE_API_KEY": "your_actual_api_key",
"KITE_SECRET_KEY": "your_actual_secret_key",
"REQUEST_TOKEN": "your_actual_request_token",
"ACCESS_TOKEN": "your_actual_access_token"
}
}
}
}
Step 3: Restart Claude Desktop
After saving the configuration, restart Claude Desktop to load the MCP server.
Alternative: Using Node.js
If you prefer Node.js over Bun:
{
"mcpServers": {
"zerodha-trading": {
"command": "node",
"args": ["--loader", "ts-node/esm", "/absolute/path/to/your/index.ts"],
"env": {
"KITE_API_KEY": "your_actual_api_key",
"KITE_SECRET_KEY": "your_actual_secret_key",
"REQUEST_TOKEN": "your_actual_request_token",
"ACCESS_TOKEN": "your_actual_access_token"
}
}
}
}
⚠️ Important:
- Replace
/absolute/path/to/your/index.tswith the actual path to your project- Use your real Zerodha API credentials
- Ensure the path uses forward slashes even on Windows
📖 API Reference
The MCP server exposes the following tools for AI interaction:
👤 Profile Operations
get-profile
Retrieves comprehensive user profile information.
// Usage in Claude
"Can you show me my Zerodha profile?"
// Returns
{
user_id: string,
user_name: string,
email: string,
user_type: string,
broker: string,
// ... additional profile data
}
💰 Portfolio & Positions
show-portfolio
Displays complete portfolio holdings with current valuations.
// Usage in Claude
"What's in my portfolio right now?"
// Returns detailed holdings with:
// - Stock symbols and quantities
// - Current market prices
// - P&L calculations
// - Total portfolio value
show-positions
Shows all open positions with live P&L.
// Usage in Claude
"Show me my current trading positions"
// Returns active positions with:
// - Entry prices and quantities
// - Current market prices
// - Unrealized P&L
// - Position status
📈 Trading Operations
buy-stock
Places a buy order for specified stock and quantity.
// Parameters
{
stock: string, // Stock symbol (e.g., "RELIANCE", "TCS")
quantity: number // Number of shares to buy
}
// Usage in Claude
"Buy 10 shares of RELIANCE"
"Purchase 50 TCS stocks"
sell-stock
Places a sell order for specified stock and quantity.
// Parameters
{
stock: string, // Stock symbol (e.g., "RELIANCE", "TCS")
quantity: number // Number of shares to sell
}
// Usage in Claude
"Sell 5 shares of HDFC Bank"
"Exit my position in Infosys"
🔍 Example Conversations
Here are some natural language examples you can use with Claude:
🗣️ "What's my current portfolio value?"
🗣️ "Buy 25 shares of ITC"
🗣️ "Show me all my losing positions"
🗣️ "What's my profile information?"
🗣️ "Sell half of my Wipro holdings"
🛠️ Development
Technology Stack
| Technology | Purpose | Version |
|---|---|---|
| TypeScript | Type-safe development | ^5.0.0 |
| KiteConnect | Zerodha API integration | Latest |
| MCP SDK | Model Context Protocol | Latest |
| Zod | Runtime type validation | ^3.0.0 |
| Bun | Fast runtime & package manager | ^1.0.0 |
Project Structure
zerodha-claude-mcp/
├── 📄 index.ts # MCP server entry point
├── 📄 trade.ts # Trading logic and API calls
├── 📄 package.json # Dependencies and scripts
├── 📄 tsconfig.json # TypeScript configuration
├── 📄 .env # Environment variables
├── 📄 README.md # Documentation
├── 📄 bun.lock # Lock file
├── 🐳 Dockerfile # Docker image configuration
├── 🐳 docker-compose.yml # Docker Compose setup
└── 📄 .dockerignore # Docker ignore rules
Local Development
-
Clone and setup (if not already done):
git clone https://github.com/gamandeepsingh/zerodha-claude-mcp.git cd zerodha-claude-mcp bun install -
Start development server:
# Development mode with auto-reload bun run dev # Production mode bun run start -
Run tests (if available):
bun test
Adding New Features
To add new trading tools:
- Define the tool in your MCP server
- Add Zod schema for parameter validation
- Implement handler using KiteConnect API
- Update documentation in this README
Debugging
Enable debug mode by setting DEBUG=true in your .env file:
DEBUG=true
This will provide detailed logging of API calls and responses.
🤝 Contributing
We welcome contributions! Here's how you can help:
Ways to Contribute
- 🐛 Bug Reports: Found a bug? Open an issue
- ✨ Feature Requests: Have an idea? Start a discussion
- 📝 Documentation: Improve docs, add examples
- 🔧 Code: Fix bugs, add features, improve performance
Development Workflow
- Fork the repository
- Create a feature branch:
git checkout -b feature/amazing-feature - Commit your changes:
git commit -m 'Add amazing feature' - Push to the branch:
git push origin feature/amazing-feature - Open a Pull Request
Code Standards
- ✅ Use TypeScript for all new code
- ✅ Follow existing code style
- ✅ Add tests for new features
- ✅ Update documentation
- ✅ Ensure all tests pass
📄 License
This project is licensed under the MIT License - see the LICENSE file for full details.
MIT License Summary
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software...
⚠️ Disclaimer
IMPORTANT: This project is for educational and development purposes only.
Risk Warning
- 📊 Trading Risk: Trading in financial markets involves substantial risk of loss
- 💰 Capital Risk: Never trade with money you cannot afford to lose
- 🔍 Due Diligence: Always conduct your own research before making trading decisions
- 🤖 AI Limitations: AI-generated trading decisions should not be followed blindly
- 📈 Market Volatility: Past performance does not guarantee future results
Best Practices
- Start Small: Test with small amounts first
- Paper Trading: Use Zerodha's paper trading feature initially
- Risk Management: Set stop-losses and position limits
- Stay Informed: Keep up with market news and regulations
- Continuous Learning: Educate yourself about trading strategies
Support
- 📚 Documentation: Check this README and inline code comments
- 💬 Discussions: Join GitHub Discussions
- 🐛 Issues: Report bugs via GitHub Issues
- 📧 Email: Contact the maintainers for specific questions
<div align="center">
Made with ❤️ for the trading community
⭐ Star this repo if you find it useful! | 🍴 Fork it to customize for your needs
</div>
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。