Python framework for building, backtesting and live-trading stock, options, futures, forex, crypto and prediction-market strategies, with optional AI agents that make trading decisions.
galleonlabs/hypergrok-trading-desk
A Hyperliquid trading desk for AI agents: seven role profiles, seventeen skills and read-only market tools, with every trade gated by your approval.
- Stars
- 366
- Forks
- 85
- Open issues
- 1
- Last commit
- 3 days ago
Repolode describes open-source projects and does not give investment advice. Trading and investing can lose money: assess the risk yourself and follow the laws where you live before using this project.
What is it?
HyperGrok is a set of instructions, role profiles and Python tools that give AI agents a shared process for Hyperliquid research, risk checks, execution and review. It defines seven roles, from a Desk Lead and Market Analyst to a Risk Manager, Execution Trader and Trade Reviewer, plus seventeen skills in the portable SKILL.md format.
The zero-key part works without a wallet: the Opening Bell script pulls a live snapshot of price, funding, open interest, spread and order book depth from Hyperliquid's public endpoint using only the Python standard library. Other scripts collect historical data into verifiable research bundles and check the desk setup.
Trading is optional and gated. Orders require a trade-only API wallet, an exact ticket that the user approves by id, and sizing based on a stressed stop. The project runs in Grok Bot and also ships guides for Claude Code, Cursor, Codex, Gemini CLI and Grok Build.
Key features
- 01Seven role profiles that split research, risk, execution and review across agents.
- 02Seventeen SKILL.md skills covering Hyperliquid setup, market data, orders and desk processes.
- 03Zero-key Opening Bell snapshot of price, funding, open interest and order book depth.
- 04Research dataset collection with hashed manifests and offline verification.
- 05Trades require explicit approval by ticket id before any order is sent.
- 06Runs in Grok Bot, Claude Code, Cursor, Codex, Gemini CLI and Grok Build.
Use cases
- Pulling a sourced market snapshot of a Hyperliquid perpetual before deciding on a trade.
- Collecting and verifying historical data for backtesting a funding-based strategy.
- Running a structured approval workflow for AI-assisted trades with a testnet wallet first.
- Giving multiple AI coding or agent tools a common trading process and checklist.
Quick start
- 1
Clone the repository at the pinned release
git clone --depth 1 --branch v1.7.1 https://github.com/galleonlabs/hypergrok-trading-desk.git - 2
Enter the project folder
cd hypergrok-trading-desk - 3
Run Opening Bell for a live ETH market snapshot (no keys needed)
python3 scripts/opening_bell.py --coin ETH
Things to know
- Full trading requires your own API wallet and verified approval controls before mainnet use.
- The full seven-Bot desk setup is designed for Grok Bot, while other runtimes differ in role and approval support.
- Perpetual futures trading can liquidate an account, and the project is documentation and instructions, not financial advice.
In collections
Topics
FAQ
- What is hypergrok-trading-desk?
- A Hyperliquid trading desk for AI agents: seven role profiles, seventeen skills and read-only market tools, with every trade gated by your approval. HyperGrok is a set of instructions, role profiles and Python tools that give AI agents a shared process for Hyperliquid research, risk checks, execution and review. It defines seven roles, from a Desk Lead and Market Analyst to a Risk Manager, Execution Trader and Trade Reviewer, plus seventeen skills in the portable SKILL.md format.
- Is hypergrok-trading-desk free to use? What license does it have?
- hypergrok-trading-desk is open source under the MIT license. Check the license terms before using it in commercial products.
- Is hypergrok-trading-desk still maintained?
- Actively maintained. The last commit to hypergrok-trading-desk was on Nov 4, 2026.
- What can I use hypergrok-trading-desk for?
- Pulling a sourced market snapshot of a Hyperliquid perpetual before deciding on a trade. Collecting and verifying historical data for backtesting a funding-based strategy. Running a structured approval workflow for AI-assisted trades with a testnet wallet first. Giving multiple AI coding or agent tools a common trading process and checklist.
- How do I install hypergrok-trading-desk?
- Clone the repository at the pinned release: git clone --depth 1 --branch v1.7.1 https://github.com/galleonlabs/hypergrok-trading-desk.git Enter the project folder: cd hypergrok-trading-desk Run Opening Bell for a live ETH market snapshot (no keys needed): python3 scripts/opening_bell.py --coin ETH
- Where can I get hypergrok-trading-desk?
- The source code and installation instructions are on GitHub: https://github.com/galleonlabs/hypergrok-trading-desk
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Open-source framework for designing, backtesting and deploying automated market-making and trading bots across many centralized and decentralized crypto exchanges.
Free open-source crypto trading bot that automates grid, DCA, AI, and TradingView strategies across Binance, Coinbase, Hyperliquid and 15+ exchanges.
An open marketplace of SKILL.md-based skills that let AI agents search tokens, trade, track wallets and use DeFi protocols on crypto exchanges and chains.
A prompt-based skill that makes AI coding agents write only the code a task needs, favoring reuse and minimal solutions over large implementations.
An open-source agent plugin for designing software: an agent drafts specs, then fresh zero-context subagents review them before any code is written.