Build a Hyperliquid bot with Claude
A bot, in Keel's sense, is a strategy that runs on its own schedule with no assistant in the loop. Claude is your copilot for building it: it composes the strategy graph, backtests it on real Hyperliquid history and helps you iterate. Four steps, one prompt each.
- 1Customize → Connectors → Add custom connector
- 2Paste https://mcp.usekeel.io/mcp and choose Connect
- 3Sign in or sign up on the Keel page that opens (no account needed beforehand)
https://mcp.usekeel.io/mcpSign in in the browser when asked. No API keys. Connect → sign in to Keel → Allow
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.
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.
Four steps from thesis to a tested strategy.
Define the thesis
Say what the bot should do in one sentence: the signal, the universe, the rebalance cadence, the sizing rule. Carry receivers versus payers sized by volatility; cross-sectional momentum on 4-hour bars. A clear thesis produces a clean graph.
Have Claude compose it
Hand Claude the thesis. It searches the 218 typed components for the signal and the regime, then writes the strategy in the Keel DSL. The graph is validated as it is composed — a wrong-shape connection fails here, not at backtest.
Backtest on real history
Ask Claude to run it. The engine reads Hyperliquid 15-minute bars and hourly carry, applies the real fee schedule and per-asset slippage, and separates the price return from the carry return. Claude summarises it in the chat.
Iterate, then review in the app
A first backtest is rarely the final bot. Ask Claude to fork it, change one parameter and compare. 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.
Questions about building a bot with Claude
What kinds of strategies can I build this way?
Anything the Keel DSL expresses: carry, cross-sectional momentum, regime-gated portfolios, volatility-targeted multi-asset systems, mean reversion, or a mix of them. If the components exist in the library, Claude can compose it.
Do I need to know Python?
No. Claude writes the strategy in the Keel DSL; your part is to read it, sanity-check the components and parameters, and decide what to keep.
How long does a build-backtest-iterate loop take?
A first backtest of a simple strategy is minutes — most of it is Claude choosing components and the backtest running. Each iteration after that is one prompt: fork, change, compare.
How do I share what I built?
Ask Claude for a share link. It creates a public page with the metrics, the equity curve and the parameters; anyone with the link can read it without an account.