Documentation

Build, test, and deploy
trading agents.

Everything from the first paper trade to production strategies. If you're brand new, start with the quickstart below.

⌘K

Getting started Strategies Exchanges & connections Paper vs live trading Backtesting Baskets

Quickstart

Your first agent, in about 5 minutes

Every step works on paper trading first — no credit card, no exchange required.

Start now

Step 1

Describe your strategy

Paste Python, Pine Script, or plain-English. No template hunting — the AI interprets any sensible phrasing.

Step 2

Review the parsed rule

We show you the canonical rule we generated. Tweak indicators, timeframes, risk controls, and allocation before it goes near an exchange.

Step 3

Connect an exchange

Binance, Coinbase, Kraken (spot + perps), Hyperliquid, KuCoin, or Trading 212. Or start on paper with a virtual $10,000.

Step 4

Deploy & monitor

Watch the agent in the dashboard: live P&L, benchmark vs BTC / S&P, proximity bars for each trigger, and a full audit log.

Example

What a real strategy looks like

You can write it exactly like this, or describe the same idea in plain English. The AI produces a canonical rule object with indicators, timeframes, entry/exit clauses, and risk controls.

Try it now Full strategy reference

strategy.rule

// Momentum breakout on BTC/USDT, Binance, 1h
// Entry:  RSI(14) crosses above 55 AND price > EMA(50)
// Exit:   Take-profit +6% OR trailing stop 3% OR RSI < 40
// Risk:   4% of portfolio per trade, max 3 concurrent positions

when RSI(14, 1h) crosses above 55 and close > EMA(50, 1h):
  buy BTC/USDT with 4% of portfolio

when RSI(14, 1h) < 40 or profit >= 6% or trailing_stop(3%):
  sell all BTC/USDT