Descriptions:
This All About AI tutorial walks through building a fully functional AI trading agent on HyperLiquid, a decentralized perpetuals platform that recently expanded beyond crypto to include instruments like S&P 500 futures and Brent crude. The video is as much an AI engineering tutorial as it is a trading one — the bulk of the runtime is spent inside Claude Code constructing the agent architecture.
The creator uses Claude Code’s custom skills system to build two discrete capabilities: a trade-discovery skill that generates intraday setups, and a research skill that evaluates individual ideas before execution. The agent is given a deliberately high-risk persona modeled after Wall Street Bets culture — complete with a stats sheet tracking “FOMO aura” and patience scores — to illustrate how persona and system prompt shape agent behavior. The full setup covers MetaMask wallet creation on Arbitrum, USDC deposit flow, HyperLiquid API key generation, and environment variable management.
Beyond the entertainment value of the WSB trader persona, the video offers a reproducible pattern for anyone wanting to pair a Claude Code agent with an external API: structured skills, a .env file for credentials, and Claude Code as the orchestrator. Developers interested in autonomous agents with real-world tool access will find the architecture transferable well outside the trading domain.
📺 Source: All About AI · Published May 20, 2026
🏷️ Format: Hands On Build
