Hyperliquid MCP Server
Enables interaction with the Hyperliquid DEX for retrieving market data, managing positions, and executing trades. Supports both testnet and mainnet operations with comprehensive trading tools including order placement, cancellation, and portfolio management.
README
Hyperliquid MCP Server
A Model Context Protocol (MCP) server for interacting with the Hyperliquid DEX. This server provides tools for retrieving market data, managing positions, and executing trades on Hyperliquid.
Built with Python 3.11+ and uv package manager, with full Docker support for easy deployment.
Features
Market Data Tools
get_all_mids- Get current mid prices for all coinsget_l2_book- Get L2 order book snapshot for a specific coinget_candle_snapshot- Get historical candle data
Account Information Tools
get_open_orders- Get all open ordersget_user_fills- Get trading history (fills)get_user_fills_by_time- Get fills for a specific time rangeget_portfolio- Get portfolio information including PnL and margin
Trading Tools
place_order- Place limit ordersplace_trigger_order- Place stop-loss or take-profit orderscancel_order- Cancel specific orderscancel_all_orders- Cancel all open orders
Installation
Prerequisites
- Python 3.11 or higher
- uv package manager
- Docker (optional, for containerized deployment)
Local Installation
-
Clone this repository
-
Install dependencies with uv:
uv sync -
Run the server:
uv run python -m hyperliquid_mcp_server.main
Docker Installation
-
Clone this repository
-
Build and run with Docker Compose:
# Production mode make build && make run # Or using docker-compose directly docker-compose up --build -
For development with hot reloading:
# Development mode make dev # Or using docker-compose directly docker-compose --profile dev up --build
Configuration
Configure the server using environment variables:
Required for Trading Operations
HYPERLIQUID_PRIVATE_KEY- Your wallet's private key (with 0x prefix)
Optional
HYPERLIQUID_WALLET_ADDRESS- Your wallet address (derived from private key if not provided)HYPERLIQUID_TESTNET- Set to "true" for testnet, "false" or unset for mainnet
Example Environment Setup
Create a .env file (not recommended for production):
HYPERLIQUID_PRIVATE_KEY=0x1234567890abcdef...
HYPERLIQUID_WALLET_ADDRESS=0xabcdef1234567890...
HYPERLIQUID_TESTNET=true
Usage
Using Make Commands
The project includes a Makefile for common operations:
# Install dependencies
make install
# Run locally (without Docker)
make local
# Build Docker image
make build
# Run in production mode
make run
# Run in development mode
make dev
# Run tests
make test
# Format code
make format
# Clean up Docker resources
make clean
Manual Commands
# Local development
uv run python -m hyperliquid_mcp_server.main
# Docker production
docker-compose up --build
# Docker development
docker-compose --profile dev up --build
With Claude Desktop
Add this server to your Claude Desktop configuration:
Local Installation
{
"mcpServers": {
"hyperliquid": {
"command": "uv",
"args": ["run", "python", "-m", "hyperliquid_mcp_server.main"],
"cwd": "/path/to/hyperliq-mcp",
"env": {
"HYPERLIQUID_PRIVATE_KEY": "0x...",
"HYPERLIQUID_TESTNET": "true"
}
}
}
}
Docker Installation
{
"mcpServers": {
"hyperliquid": {
"command": "docker",
"args": ["run", "--rm", "-i", "--env-file", ".env", "hyperliquid-mcp-server:latest"],
"cwd": "/path/to/hyperliq-mcp"
}
}
}
API Reference
Market Data
get_all_mids
Get current mid prices for all coins.
{}
get_l2_book
Get L2 order book for a specific coin.
{
"coin": "BTC",
"nSigFigs": 3
}
get_candle_snapshot
Get historical candle data.
{
"coin": "BTC",
"interval": "1h",
"startTime": 1640995200000,
"endTime": 1641081600000
}
Account Information
get_open_orders
Get open orders for the configured wallet or a specific user.
{
"user": "0x..." // optional
}
get_user_fills
Get trading history.
{
"user": "0x..." // optional
}
get_portfolio
Get portfolio information.
{
"user": "0x..." // optional
}
Trading
place_order
Place a limit order.
{
"assetIndex": 0,
"isBuy": true,
"price": "50000",
"size": "0.1",
"timeInForce": "Gtc",
"reduceOnly": false,
"clientOrderId": "my-order-1"
}
place_trigger_order
Place a trigger order (stop-loss/take-profit).
{
"assetIndex": 0,
"isBuy": false,
"size": "0.1",
"triggerPrice": "45000",
"isMarket": true,
"triggerType": "sl",
"reduceOnly": true
}
cancel_order
Cancel a specific order.
{
"assetIndex": 0,
"orderId": 12345
}
cancel_all_orders
Cancel all open orders.
{}
Security Notes
- Never share your private key
- Use testnet for development and testing
- Consider using environment variables or secure secret management for production
- This server requires your private key to sign trading transactions
- Read-only operations (market data, account info) work without a private key
Asset Indices
Common asset indices for Hyperliquid:
- BTC: 0
- ETH: 1
- SOL: 2
- (Check Hyperliquid documentation for complete list)
Error Handling
The server includes comprehensive error handling:
- Invalid configurations are reported on startup
- API errors are caught and returned with descriptive messages
- Network timeouts are handled gracefully
- Input validation prevents malformed requests
Development
Project Structure
hyperliquid_mcp_server/
├── main.py # Main MCP server
├── types/
│ └── hyperliquid.py # Pydantic type definitions
├── utils/
│ ├── hyperliquid_client.py # API client
│ └── config.py # Configuration management
└── tools/
├── market_data.py # Market data tools
├── account_info.py # Account information tools
└── trading.py # Trading tools
Development Setup
# Install dependencies
uv sync
# Install with development dependencies
uv sync --extra dev
# Run tests
uv run pytest
# Format code
uv run black .
uv run isort .
# Lint code
uv run ruff check .
uv run mypy .
Docker Development
# Development mode with hot reloading
make dev
# Shell into container
make shell
# View logs
make logs
License
MIT
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。