Claude × Hyperliquid

Claude on Hyperliquid: build and backtest strategies in chat

Connect Claude to Keel and it becomes a Hyperliquid strategy copilot: it searches the typed component library, composes a strategy in the Keel DSL, backtests it on real Hyperliquid history and walks you through the result — on web, desktop, phone or in Claude Code.

Or set it up directly
Claude
  1. 1Customize → Connectors → Add custom connector
  2. 2Paste https://mcp.usekeel.io/mcp and choose Connect
  3. 3Sign in or sign up on the Keel page that opens (no account needed beforehand)
Keel MCP endpoint
https://mcp.usekeel.io/mcp

Sign in in the browser when asked. No API keys. Connect → sign in to Keel → Allow

Check it worked

Two ways to tell, and either one settles it — a connection that is already live is not improved by signing in again.

In Claude Keel is listed where you added it (Customize → Connectors → Add custom connector) once the sign-in has finished.

Ask the agent keel_account_status answers with the Keel account the session is signed in as, which nothing but a live connection can return. If the agent says it has no Keel tools, the sign-in did not finish.

Prompt
Use keel_account_status and tell me which Keel account this session is signed in as.

Not yet verified against a Claude version. These steps follow Anthropic's own documentation, read on 2026-09-16.

Ask AI about Keel
Other ways to connect
Every client, step by step
What Claude does

Propose, compose, backtest, review.

  1. Propose a thesis

    Describe the idea in plain language: cross-sectional momentum on the top 30 perps, a carry strategy sized by volatility, a regime-gated mix. Claude turns it into a component graph.

  2. Compose in the Keel DSL

    Claude reads the 218 typed components and writes the strategy against them. The graph is validated as it is composed, so a bad connection fails before any backtest.

  3. Backtest on real history

    One request runs it on Hyperliquid 15-minute bars and hourly carry with real fees and slippage. Claude summarises Sharpe, drawdown and exposure, and forks a variant when you ask.

  4. Review in the app

    The agent builds, tests and reads. When you want to run a strategy on your account, it hands it to the Keel app and you take it from there.

FAQ

Questions about Claude and Hyperliquid

How do I connect Claude to Hyperliquid through Keel?

Copy the prompt above into Claude and follow its steps, or add the Keel MCP server as a connector: the box above has the URL and the steps for Claude on web, desktop and phone. Sign in in the browser when asked. No API keys.

Does this work in Claude Code and Claude Desktop too?

Yes. Claude Desktop uses the same connector as the web. Claude Code adds the server with one command and signs in through /mcp — the Claude Code guide on the agents page walks through it.

How is this different from asking Claude to write a Python script?

A script that calls the exchange API is not an engine: no portfolio simulation, no fee schedule, no carry accounting, no version history. Through Keel, Claude composes against a typed component graph that the engine validates, backtests and keeps under version control.

What does Claude actually call?

The 29 tools the Keel MCP server publishes, by name: component search and detail, compose, backtest and summarise, fork and diff, the strategy library, share links and a read of the result. The full list, grouped, is on the agents page.