Sign In with Google

AlphaGen-AI uses Google OAuth for secure authentication. Click "Login" from any page and sign in with your Google account. Your account is automatically created on first login.

Plan Overview

New accounts start in Preview mode (read-only). Contact an administrator or upgrade to access features:

Key Locations

Generating a Single-Model Idea

1. Navigate to /dashboard
2. Select your preferred AI model from the dropdown (GPT-4o, Gemini 2.5/3, Claude, DeepSeek, GLM, Qwen, Grok)
3. (Optional) Upload a BTC chart screenshot (PNG/JPG, max 5 MB) for visual pattern analysis
4. Click "Generate Trade" — the AI analyzes live CoinGlass derivatives data
5. Review the generated trade idea with entry, stop loss, take profit, and confidence levels

Available AI Models

Trade Idea Output

Every trade idea includes:

8-Point Signal Validation Framework

Every AI trade idea must validate against these 8 signals with a weighted score ≥40%:

Trade History

All saved trade ideas appear in the Trade History panel on the dashboard. Track your P&L, filter by date range, and analyze model performance over time. Starter plans retain 30 days, Plus retains 90 days, Pro has unlimited history.

Data Sources

The Market Rotation Dashboard pulls live data from 11 CoinGlass API endpoints:

• • Funding Rate (OI-weighted)
• • Open Interest Aggregated
• • Liquidation Data
• • ETF Flow Tracker
• • Orderbook Depth
• • Taker Buy/Sell Volume

6-Signal Composite Scoring

The dashboard calculates a composite rotation score from 6 weighted signals:

1. Funding Spread — BTC vs altcoin funding differential
2. OI Divergence — Open interest flow between BTC and alts
3. Volume Profile — Trading volume concentration
4. BTC Dominance — Market cap share trend
5. Alt Season Index — Alternative.me alt season indicator
6. Sector Rotation — Capital flow between crypto sectors

Rotation Levels

Data Sources

The Social Sentiment Gauge combines three data sources via the /api/context/sentiment endpoint:

• Fear & Greed Index (Alternative.me) — 0-100 scale from "Extreme Fear" to "Extreme Greed"
• CoinGlass Funding Rate — 8-hour OI-weighted funding with historical percentile
• CoinGlass Long/Short Ratio — Aggregate trader positioning across exchanges

Composite Weighting

Example: Fear & Greed at 28 (Fear) + negative funding + net shorts = "Moderate Long" bias with contrarian signal.

Requirements

• • Hyperliquid Wallet — Create at app.hyperliquid.xyz
• • API/Agent Key — Generate in Hyperliquid settings (we encrypt it with MASTER_KEY)
• • TradingView Account — With alert webhooks enabled (Pro/Premium plan)
• • AlphaGen-AI Pro Plan — Required for bot access

Step 1: Connect Hyperliquid Credentials

1. Navigate to /dashboard/hyperliquid
2. Click "Connect Wallet" in the credentials section
3. Enter your Hyperliquid wallet address and API/agent private key
4. Keys are encrypted with AES-256-GCM and stored securely

Step 2: Configure Bot Settings

• • Leverage: Set between 1x-50x (default: 10x)
• • Position Sizing: Fixed USD or percentage of balance
• • Webhook Passphrase: Create a secure secret (required for webhook auth)
• • Kill Switch: Enable to start accepting trades

Step 3: Configure TradingView Webhook

Webhook URL:

Alert Message Format:

{
  "passphrase": "YOUR_WEBHOOK_SECRET",
  "ticker": "{{ticker}}",
  "side": "{{strategy.order.action}}",
  "size": "{{strategy.order.contracts}}"
}

In TradingView alert settings, set the message to: {{strategy.order.alert_message}}

Step 4: Verify & Monitor

• • Test Signal Button: Send a test webhook to verify connectivity
• • Bot Logs Section: View incoming alerts and Hyperliquid exchange responses
• • Position Monitor: Real-time P&L tracking for open positions

Security Features

4-layer OWASP Top 10 security hardening:

• • Global Auth Guard: Path traversal prevention with dot segment resolution
• • Helmet CSP: Production-safe Content Security Policy
• • Rate Limiting: 10 req/hour for webhooks, 1200 req/hour for polling
• • API Sanitization: Input validation and request sanitization

Setup Steps

Creating a DCA Strategy

You can create DCA strategies from two sources:

DCA Parameters

Ladder Preview

Before activating, the system generates a preview showing:

• • Safety Order Levels: Exact trigger prices for each order
• • Order Sizes: USD amount at each level
• • Average Entry: Projected average after all orders fill
• • Liquidation Buffer: Distance to liquidation price at each level

Whale Positions Panel

Track the top 20 BTC positions exceeding $1M from CoinGlass and Hyperliquid data:

• • Position Size: Total USD value of each whale position
• • Direction: Long or short bias
• • Entry Price: Estimated average entry
• • Aggregate Bias: Net long/short ratio across all whales

ETF Flow Tracker

Monitor Bitcoin ETF capital flows:

• • Daily Flows: Net inflows/outflows per day
• • Weekly Aggregates: 7-day cumulative flow
• • Threshold Alerts: Notifications when flows exceed $500M

Kelly Criterion Calculator

Optimal position sizing based on your edge:

• • Inputs: Win rate (%), average reward/risk ratio
• • Full Kelly: Theoretically optimal bet size
• • Fractional Kelly (0.25-0.5x): Recommended conservative sizing

Liquidation Heatmap

Visual representation of liquidation clusters at various price levels. Use this to identify potential stop-hunting zones and optimal stop-loss placement outside of major liquidation clusters.

How Limits Are Enforced

• • Ideas per Day: Counter resets at midnight UTC. Rejected ideas don't count.
• • Model Access: Only available models appear in the dropdown for your plan.
• • Bot Access: Hyperliquid dashboard locked for non-Pro users.
• • Feature Gates: UI elements for locked features show upgrade prompts.
• • Pro TradingView Strategy: Includes the official AlphaGen-AI TradingView Strategy (Pine Script) for automated BUY/SELL signals that connect your TradingView charts to the Hyperliquid bot via webhooks.

AI-Enforced Risk Controls

• Minimum Risk/Reward (1.5:1):

  Every AI trade idea must have a risk/reward ratio of at least 1.5. Ideas failing this check are rejected and don't count toward your daily limit.

• Signal Score Threshold (40%):

  The weighted 8-point signal score must exceed 40% for the idea to be approved.

• Confidence Calibration:

  Each model has bias correction applied to normalize confidence scores across different AI reasoning approaches.

Confidence Ranges by Signal Count