Skip to main content
Every platform has its own vocabulary. This page is the short, plain-English dictionary. When something in the app confuses you, look here first.

Trading basics

Perpetual contract. A crypto derivative that tracks an underlying asset (like BTC) without an expiry date. You post margin, open a long or short position, and settle any profit or loss when you close. Long / short. Long means you profit when price goes up. Short means you profit when price goes down. Leverage. A multiplier on your position size relative to the margin you posted. 10x leverage means a 1,000positionwith1,000 position with 100 margin. Higher leverage means bigger wins and bigger losses. Liquidation. When your losses eat through your margin, the exchange force-closes your position. Higher leverage makes this much easier to trigger. Liquidations are the single biggest way traders lose money on perpetuals. Take profit / stop loss. Automatic exit orders. Take profit closes your position when price reaches a target. Stop loss closes it when price moves against you by a set amount. Hyperoru attaches these to decisions whenever the strategy specifies them. Funding rate. A periodic payment between longs and shorts that keeps the perpetual price close to the underlying spot price. Positive funding means longs pay shorts; negative means the opposite. Open interest. The total number of contracts currently held open on an exchange. Rising open interest plus rising price usually means a real trend; rising open interest plus flat price often precedes a volatility spike.

Hyperoru concepts

AI trader. A standalone trading account inside Hyperoru. Each trader has its own name, its own language-model setup, its own exchange wallet, and its own strategy. You can run many in parallel. Strategy. The logic that decides what to do when the market is analyzed. Hyperoru supports two kinds:
  • Prompt strategy — natural-language instructions interpreted by a language model.
  • Program strategy — a Python class with a should_trade(data) method.
Trading prompt. The template your prompt strategy uses. It combines your instructions with live market context (prices, positions, indicators, news) before the language model reads it. Trading program. The Python class a program strategy uses. Runs in a sandbox — no network, no filesystem — so it cannot do anything other than return a decision. Signal pool. The logic that decides when a strategy should run. It groups signal conditions with a set of symbols and fires when the conditions match. Strategies only run after a signal fires. Signal definition. A single market condition — for example “BTC open interest surges more than 10% in 5 minutes.” You combine definitions with AND/OR into a signal pool. Binding. The link between a program strategy and a trader. A binding includes the program, its signal-pool triggers, a schedule, and exchange-specific settings. Decision. The output of a strategy run. Contains the operation (buy, sell, close, hold), the symbol, the position size, leverage, take-profit/stop-loss prices, and a plain-language reason. Arena. The live view where you watch decisions happen. Useful for debugging: you see the exact prompt, the model’s reasoning, and the decision side-by-side. HyperAI. The conversational assistant built into Hyperoru. It can answer questions about your traders, help you build strategies, and run analytical skills like trade attribution and backtest interpretation. Factor. A computed market indicator (momentum, mean-reversion, carry, volatility, etc.). Hyperoru ranks symbols by factor score so strategies can focus on the most attractive names. Market regime. A classification of the current market: breakout, trending, ranging, volatile, or quiet. Strategies can gate their behavior on the current regime. Backtest. Running a strategy against historical data to see how it would have performed. Does not place real trades. Session token. The long string you attach to API requests to prove you are logged in. Treat it like a password. See Authentication.

API concepts

REST. The main HTTP API. You send JSON over HTTPS; the server returns JSON. WebSocket. A bidirectional connection for live updates. You get prices, orders, positions, and AI decisions pushed to you instead of polling. See WebSocket. Rate limit. A cap on how many requests you can send per minute. See Errors and rate limits. Idempotency. The property that doing the same operation twice is the same as doing it once. Critical for mutating calls. If the Hyperoru API ever retries an internal operation you will not see double-trades.

Risk terms

Drawdown. The drop from a previous high-water mark in your equity curve. A 20% drawdown means you are down 20% from your best equity value. Sharpe ratio. Return per unit of risk. Higher is better. Positive Sharpe means returns beat risk-free rate; negative Sharpe means you would be better off in cash. Win rate. The fraction of trades that are profitable. High win rate alone does not mean profitable — a strategy with 90% win rate and 10:1 loser-to-winner ratio still loses money. Expected value (EV). Average PnL per trade accounting for win rate and win/loss sizes. This is the number that matters.

Still confused?

Search this site, ask HyperAI inside the app, or email support@hyperoru.com. If a term on these docs is unclear, email us and we will add it here.