AI Automated Trading Bot

Linux-only automated trading system for Solana blockchain with arbitrage detection, liquidation monitoring, and risk management.

⚠️ Important: This bot is designed and tested for Linux environments only. It may not work properly on macOS or Windows.

What is MEV?

MEV (Maximal Extractable Value) is a trading strategy that exploits price differences across multiple DEXes on Solana testnet. The bot executes multi-hop trades like:

SOL → Coin1 → Coin2 → Coin3 → SOL

This creates arbitrage opportunities by taking advantage of price inefficiencies between different decentralized exchanges.

Features

• MEV Trading: Multi-hop arbitrage (SOL → Coin1 → Coin2 → Coin3 → SOL)
• AI Trading: OpenAI integration for market analysis
• Arbitrage Bot: Multi-DEX support with path discovery
• Liquidation Bot: Solend protocol integration
• Risk Management: Built-in risk assessment and limits
• Testnet Only: Safe testing environment on Solana testnet
• Real-time Monitoring: Continuous price monitoring and opportunity detection

Project Structure

bot/
├── src/
│   ├── arbitrage-bot/          # Arbitrage detection
│   ├── liquidation-bot/        # Liquidation monitoring
│   └── utils/                  # Utility functions
├── models/                     # AI models
├── config.toml                 # Configuration
└── package.json                # Dependencies

Installation

Prerequisites

• Operating System: Linux (Ubuntu 18.04+, CentOS 7+, or similar)
• Node.js (v16+)
• Solana CLI tools
• Anchor Framework

Setup

1. Clone and install:

git clone https://github.com/xodn348/AI_automated_trading.git
cd AI_automated_trading
npm install

2. Configure environment:

# Create .env file in src/arbitrage-bot/
OPENAI_API_KEY=your_openai_api_key
SOLANA_PRIVATE_KEY=your_solana_private_key
SOLANA_RPC_URL=https://api.devnet.solana.com

3. Important: Fix wallet path issue

# Copy wallet.json to src/ directory (required for bot to start)
cp wallet.json src/wallet.json

# Or modify config.toml to use correct path
# Comment out: private_key_path = "/home/jnnj92/bot/wallet.json"

Linux Deployment

Quick Setup

# Make deployment script executable
chmod +x deploy-linux.sh

# Run deployment script
./deploy-linux.sh

Docker Deployment

# Build and run with Docker
docker-compose up -d

# Or build manually
docker build -t ai-trading-bot .
docker run -d --name ai-trading-bot ai-trading-bot

Systemd Service

# Start as system service
sudo systemctl start ai-trading-bot

# Enable auto-start
sudo systemctl enable ai-trading-bot

# Check status
sudo systemctl status ai-trading-bot

# View logs
sudo journalctl -u ai-trading-bot -f

Usage

Start Arbitrage Bot

# Main bot (requires wallet.json in src/ directory)
npm start

# AI-enhanced version
node src/arbitrage-bot/index2_ai.js

# Test bot startup (recommended first)
node -e "try { require('./src/arbitrage-bot/index.js'); console.log('Bot loaded successfully!'); } catch(e) { console.error('Error:', e.message); }"

Important Notes

• Wallet Setup: Ensure wallet.json exists in src/ directory
• Devnet Only: Bot currently works only on Solana devnet for safety
• Test Mode: Bot runs in monitoring mode without executing real trades
• API Keys: OpenAI API key required for AI features to work

Start Liquidation Bot

# Liquidation monitoring
node src/liquidation-bot/liquidation-bot.js

# Create test positions
node src/liquidation-bot/create-test-position.js

Testing

cd test-liquidation
npm test

Configuration

Edit config.toml for:

• Network settings (RPC endpoints)
• Trading parameters (slippage, position sizes)
• AI settings (OpenAI configuration)
• Risk management (stop-loss, take-profit)

Current Settings

• Network: Solana testnet only (https://api.testnet.solana.com)
• Trading: Test mode (no real transactions executed)
• MEV Strategy: Multi-hop arbitrage across multiple DEXes
• Monitoring: 30-second intervals for arbitrage opportunities
• Safety: 1% minimum profit threshold before execution

Security

• API keys stored in .env (not committed to Git)
• Configurable trading limits
• Oracle price validation
• Slippage protection

Data Collection

Automatically collects:

• Historical price data
• Trading performance metrics
• Market volatility data
• Arbitrage opportunity logs

License

MIT License - see LICENSE file for details.

Disclaimer

For educational and research purposes only. Cryptocurrency trading involves significant risk. Use at your own risk.