System Overview
This page helps you understand how Hyper Alpha Arena works from end to end.
How the System Works
{BTC_PRICE}{BTC_RSI14_15m}{BTC_MACD_1h}{BTC_EMA20_4h}{BTC_CVD_1h}{BTC_OI_CHANGE}{BTC_FUNDING_RATE}{trigger_market_regime}{trigger_context}Step-by-Step Breakdown
1. Market Data Collection (Automatic)
You don't need to do anything here - the system runs 24/7 automatically.
The system continuously collects data from Hyperliquid:
| Data Type | Details | Purpose |
|---|---|---|
| K-Line Data | Multiple timeframes (1m/5m/15m/1h/4h/1d) | Price action analysis |
| Technical Indicators | RSI, MACD, EMA, Bollinger Bands, etc. | Trend & momentum |
| Market Flow | CVD, OI, Funding Rate, Taker Volume | Money flow analysis |
| Order Book | Depth Ratio, Order Imbalance | Buy/sell pressure |
| Market Regime | breakout/absorption/exhaustion, etc. | Market state classification |
Understanding Market Regime
Market Regime is the system's automatic classification of current market conditions:
- breakout: Price breaking key levels, potential trend start
- absorption: Large orders being absorbed, potential reversal
- exhaustion: Trend momentum fading, potential top/bottom
The AI uses this to determine what actions are appropriate for current conditions.
2. Signal Detection
Think of Signal Pool as an "alarm clock" - it doesn't decide what you do after waking up, it just wakes up the AI when certain conditions are met.
| Feature | Page | Description |
|---|---|---|
| Signal Pool | Signal Pools | Define trigger conditions |
| AI Signal Generator | AI Signal Generator | Let AI help create configurations |
Here's an example:
- You set up a signal:
RSI < 30 - When BTC's RSI drops below 30, the signal fires
- AI Trader "wakes up" and starts analyzing whether to trade
- But the AI might decide NOT to buy! It also considers other indicators, Market Regime, and your Prompt rules
Key Concept
Signal Pool only defines trigger conditions - it does NOT decide trade direction.
Many users mistakenly think setting "RSI < 30" tells the AI to go long. That's incorrect. After being triggered, the AI analyzes everything and may go long, short, or do nothing.
3. Prompt Configuration
Trading Prompt is the AI's "trading playbook" - it defines the AI's trading style, risk preferences, and decision rules.
| Feature | Page | Description |
|---|---|---|
| Trading Prompt | Prompts & Strategies | Define trading rules |
| AI Prompt Writer | AI Prompt Writer | Let AI help craft prompts |
What can you write in a Prompt?
You are a conservative trend trader.
Entry Rules:
- Only consider entries when {trigger_market_regime} is breakout
- RSI ({BTC_RSI14_15m}) should be between 40-60
- Must have clear trend direction
Risk Management:
- Maximum 10% position size per trade
- Stop loss within 2% of entry priceThe Power of Prompts
Same signal trigger, different Prompts = completely different trading decisions:
- Aggressive Prompt: Might go all-in immediately
- Conservative Prompt: Might only open a small test position
- Cautious Prompt: Might decide not to trade, wait for better setup
4. AI Trader Setup & Triggers
Before the AI starts working, you need to complete two prerequisites:
| Setting | Description |
|---|---|
| Choose AI Model | Claude / GPT / DeepSeek - each has different characteristics |
| Connect Wallet | Enter your Hyperliquid API Key so AI can execute trades |
Once configured, there are two ways to trigger the AI:
| Trigger Method | Best For |
|---|---|
| Signal Trigger | Fires when Signal Pool conditions are met - good for catching specific opportunities |
| Scheduled Trigger | Auto-fires every N minutes - good for continuous market monitoring |
Both methods can be enabled simultaneously. See AI Trader Setup for details.
4.5 Program Trading (Optional)
Besides AI Trader, you can also use Program Trading to execute strategies.
| Comparison | AI Trader | Program Trading |
|---|---|---|
| Decision | AI model analyzes and decides | Code logic executes directly |
| Predictability | Low, AI may judge differently | High, same conditions = same results |
| Speed | Slower (waits for AI response) | Fast (direct code execution) |
When to use Program Trading?
- You have clear trading rules (e.g., buy when RSI < 30)
- You want fast, predictable strategy execution
- You don't need AI's "subjective judgment"
See Program Trading for detailed usage.
5. AI Trader Decision
When triggered, the AI reads the following information:
| Information | Source | Example |
|---|---|---|
| Trading Prompt | Your configuration | Trading style, rules, risk preferences |
| Real-time Indicators | Automatic | {BTC_RSI14_15m} = 45.2 |
| Market Regime | Automatic | {trigger_market_regime} = breakout |
| Trigger Context | Automatic | What signal fired, indicator values at that moment |
Variables you can use in your Prompt (examples):
| Category | Variable Examples |
|---|---|
| Price | {BTC_PRICE}, {ETH_PRICE} |
| Technical Indicators | {BTC_RSI14_15m}, {BTC_MACD_1h}, {BTC_EMA20_4h} |
| Market Flow | {BTC_CVD_1h}, {BTC_OI_CHANGE}, {BTC_FUNDING_RATE} |
| Market State | {trigger_market_regime}, {trigger_context} |
See Prompts & Strategies for the complete variable list.
6. ⭐ Execute Trade
This is the final step where everything comes together.
After comprehensive analysis, the AI executes decisions via Hyperliquid API:
| Action | Description |
|---|---|
| Buy | Open long or add to position |
| Sell | Open short or reduce position |
| Close | Close existing position |
| Adjust Size | Increase or decrease position size |
| Hold | Decide not to trade, wait for better opportunity |
No Manual Configuration Needed
This step is fully automatic. You just need to ensure:
- Wallet is connected with sufficient balance
- API Key has correct permissions
7. View Results
| Feature | Location | Description |
|---|---|---|
| Equity Curve | Dashboard | Real-time account value monitoring |
| AI Decisions | Inside Dashboard | See AI's reasoning for each decision |
| Attribution | Inside Dashboard | Understand what triggered each trade |
AI Decisions is your debugging tool - you can see exactly what the AI was "thinking" during each decision, helping you optimize your Prompt and Signal Pool.