Documentation
GOAT is an advanced AI-powered cryptocurrency trading platform that leverages multiple state-of-the-art language models to analyze market data and execute trades with precision.
Multi-Model AI
6 AI models working simultaneously for optimal trading decisions
Real-Time Analysis
Live market data processing every 30 seconds
Risk Management
Advanced position sizing and stop-loss strategies
Key Features
AI-Powered Trading
Our platform integrates six cutting-edge AI models that analyze market conditions from multiple perspectives:
- Claude Sonnet 4.5 - Advanced reasoning and market analysis
- DeepSeek V3.2 - Deep learning pattern recognition
- Gemini 2.5 PRO - Multi-modal market intelligence
- Grok 4 - Real-time sentiment analysis
- GPT-5 - Contextual understanding and prediction
- Qwen 3-Max - High-frequency trading signals
Multi-Timeframe Analysis
Trading decisions are based on comprehensive timeframe analysis:
- 1-Minute Charts - Precise entry timing
- 5-Minute Charts - Primary scalping analysis
- 15-Minute Charts - Standard momentum plays
- 1-Hour Charts - Trend context and major S/R levels
- 4-Hour Charts - Overall market direction
Adaptive Strategy
The system automatically adjusts trading parameters based on:
- Market volatility (measured via 5M candle ranges)
- Multi-timeframe confluence strength
- BTC correlation impact on altcoins
- Position exposure limits (max 65-70%)
AI Models
Claude Sonnet 4.5
Anthropic's flagship model excels at complex reasoning and multi-step market analysis. Provides detailed trade justifications with high confidence signals.
DeepSeek V3.2
Advanced pattern recognition model that identifies complex chart formations and historical price action similarities.
Gemini 2.5 PRO
Google's multi-modal AI processes market data, news sentiment, and social signals to generate comprehensive trading insights.
Grok 4
xAI's model with real-time data access provides up-to-the-second market sentiment and breaking news impact analysis.
GPT-5
OpenAI's latest model brings superior contextual understanding and predictive capabilities for market movements.
Qwen 3-Max
Alibaba's high-performance model optimized for rapid decision-making and high-frequency trading signal generation.
Trading System
Signal Types
LONG
Buy signal with confidence ≥ 0.6
SHORT
Sell signal with confidence ≥ 0.6
HOLD
Maintain existing position
WAIT
No clear setup, stay on sidelines
CLOSE
Exit position (invalidation or low confidence)
Risk Management Formula
Step 1: Base Risk % (5-15%)
Confidence Range Base Risk % Typical Leverage
0.8 - 1.0 13-15% 5x-10x
0.7 - 0.8 10-12% 10x-12x
0.65 - 0.7 7-9% 12x-15x
0.6 - 0.65 5-7% 15x-20x
Step 2: Calculate Base Risk
risk_usd = total_balance × base_risk%
Step 3: Apply Leverage
size_usd = risk_usd × leverage
Constraint:
size_usd / total_balance ≤ 0.65 (65% max exposure)
Trade Duration Logic
- 15-MIN: Quick scalps on 5M setups with tight confluence (15x-20x leverage)
- 30-MIN: Standard momentum plays on 15M patterns (10x-15x leverage)
- 60-MIN: Strong trend continuation on 15M/1H alignment (8x-12x leverage)
- 2-8H: Major trend plays with 1H/4H confirmation (5x-10x leverage)
Stop Loss & Take Profit
Stop Loss Placement:
- 15M trades → below/above nearest 1M or 5M swing
- 30M trades → below/above 5M or 15M S/R
- 60M trades → below/above 15M or 1H S/R
- Min distance: 0.3% from entry
- Adaptive based on volatility (0.3%-2.5%)
Take Profit:
- LONG → nearest resistance level
- SHORT → nearest support level
- Uses Fibonacci retracements + psychological levels
- Min distance: 0.5% from entry
- Minimum Risk:Reward ratio of 1.5:1
Setup & Configuration
Requirements
- Python 3.9 or higher
pipfor package management- MySQL/MariaDB database
- API keys for AI model providers (Claude, OpenAI, etc.)
- Aster API credentials for live trading
- HTTPS enabled domain
Installation
# Clone repository
git clone https://github.com/yourusername/goat.git
# Install Python dependencies
pip install -r requirements.txt
# Configure database (if using a new DB)
# python database.py
# Set up configuration
copy .env.example .env
# Edit configuration file with your API keys
# Open .env and add your keys
# (Permissions are typically not required on Windows)
Configuration
Create a .env file in the root directory and add your settings:
# Site Configuration
SITE_ENABLED=true
GITHUB_LINK="https://github.com/yourusername/goat"
TWITTER_LINK="https://twitter.com/yourhandle"
# AI Model API Keys
CLAUDE_API_KEY="sk-ant-api03-..."
OPENAI_API_KEY="sk-..."
ANTHROPIC_VERSION="2023-06-01"
# Trading Configuration
BINANCE_API_KEY="your_binance_key"
BINANCE_API_SECRET="your_binance_secret"
MAX_POSITION_EXPOSURE=0.65
UPDATE_INTERVAL=30
Running the System
# Start the trading bot
python aiagent/trading_bot.py
# Run in background (Windows)
start /b python aiagent/trading_bot.py > logs/bot.log 2>&1
# Monitor logs (PowerShell)
Get-Content logs/bot.log -Wait