dakkshin/hyperliquid-mcp — explained in plain English
Analysis updated 2026-05-18
Let an AI assistant place and manage cryptocurrency perpetual futures trades on Hyperliquid.
Get server side computed market signals like order book imbalance and technical indicators without pulling raw data.
Run a Monte Carlo simulation to estimate potential downside risk of holding a position over time.
| dakkshin/hyperliquid-mcp | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | — | easy | hard |
| Complexity | — | 2/5 | 5/5 |
| Audience | developer | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Hyperliquid MCP is a server that lets an AI assistant such as Claude trade on Hyperliquid, a platform for perpetual futures trading in cryptocurrency, through a standard connection method called the Model Context Protocol. In plain terms, it gives an AI model a set of tools it can call to place trades, check prices, and analyze market data on your behalf, using a wallet key that you provide. Setting it up mostly means adding a small configuration block to an AI client, pointing it at this server and supplying a private key along with a setting for whether to use the real exchange or its practice testnet version. The project strongly recommends starting on the testnet, which is a completely separate practice environment with its own funds and its own numbering for assets, before ever touching real money. For real trading, it recommends a safer setup where a separate limited permission key does the actual trading while the user's main account key never touches the machine running the AI model. The server offers around thirty tools covering order placement, canceling and modifying orders, and combined take profit and stop loss trades that are linked together so that filling one automatically cancels the other. It is built to reject clearly invalid orders before they are ever sent to the exchange, such as a stop loss placed on the wrong side of the market, and it reports honestly when something fails or times out rather than assuming success. Beyond trading, it also provides live market data tools, including order book imbalance measures, buy and sell pressure signals from recent trades, common technical indicators like RSI and moving averages, and even a simulation tool that estimates potential gains or losses over time using repeated random sampling. These calculations happen on the server itself so the AI model receives short, ready to use numbers instead of large amounts of raw market data. The project is written in Python and is still adding some features, such as time weighted average price orders, which are declared but not yet functional.
An MCP server that lets AI assistants like Claude trade on Hyperliquid perpetual futures, place bracket orders, and pull live market microstructure and risk analytics through simple tool calls.
Mainly Python. The stack also includes Python, MCP, Hyperliquid SDK.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.