ProjectSeagull

Algorithmic trading platform for backtesting and executing orders from trading strategies

📚 Documentation

→ Complete Setup & Usage Guide - Everything you need to know!

Quick Links

• Database Setup
• Configuration GUI
• Running Backtests

Database Setup

Quick Start

# 1. Install dependencies
pip install -r requirements.txt

# 2. Set database connection
$env:DATABASE_URL = "postgresql://user:pass@localhost:5432/seagull"
$env:MASSIVE_API_KEY = "your_polygon_api_key"
$env:NASDAQ_DATA_LINK_API_KEY = "your_nasdaq_key"

# 3. Initialize database
python Scripts/init_db.py

What this does:

• ✅ Creates database schema (signals, agents, tests, jobs)
• ✅ Seeds default data
• ✅ Auto-registers all agents from Agents/instances/
• ✅ Uploads agent code to database
• ✅ Ready to run backtests!

Verification:

python Common/check_agents_registry.py

See SETUP_GUIDE.md for detailed instructions.

Configuration GUI

Launch the all-in-one configuration interface:

python Scripts/general_config_gui.py
# OR: Scripts\launch_config_gui.bat

Four Modules:

1. 📡 Signal Matrix - Register trading signals from Massive/SF1
2. 🧪 Test Protocols - Create backtest configurations
3. ⚙️ Job Scheduler - Assign agents to tests
4. 🤖 Agent Factory - Register & clone agents

Key Features:

• ✅ Auto-validation of symbols and signals
• ✅ Clone agents with signal substitution
• ✅ Create agent variants in 2 minutes
• ✅ Visual browsing of agents, tests, and jobs

Example workflows:

• Register new signals → Create test → Assign agent → Run backtest
• Clone agent for new symbol → Test on multiple timeframes
• Build agent portfolio (AAPL, TSLA, MSFT variants)

See SETUP_GUIDE.md for detailed usage instructions.

Running Backtests

# Run all configured test jobs
python Scripts/run_backtest.py

# Run specific tests
$env:BACKTEST_TEST_NAMES="my_test,quick"
python Scripts/run_backtest.py

What happens:

1. Loads test definitions from database
2. Loads test jobs (agent-test pairs)
3. Loads agents from database (code column)
4. Runs backtests with configured parameters
5. Generates reports and plots

View results:

• Logs in logs/ directory
• Plots in configured plot directory
• JSON decision logs (if agent implements logging)

See SETUP_GUIDE.md for complete workflow and troubleshooting.

---

Project Structure

ProjectSeagull/
├── Agents/instances/      # Agent Python files
├── Backtesting/          # Backtest engine
├── Common/               # Shared utilities, DB access
├── Scripts/              # Tools (GUI, init, visualize)
├── db/initialize.sql    # Database schema
├── README.md            # This file
└── SETUP_GUIDE.md       # Complete documentation

Quick Reference

# Setup
python Scripts/init_db.py                    # Initialize database
python Common/check_agents_registry.py       # Verify setup

# Configure
python Scripts/general_config_gui.py         # Launch GUI

# Test
python Scripts/run_backtest.py              # Run backtests

# Monitor
python Scripts/view_signal_usage.py         # Signal usage stats

Getting Help

• Complete guide: SETUP_GUIDE.md
• Hybrid agent template: Agents/instances/HYBRID_AGENT_GUIDE.md
• Troubleshooting: See SETUP_GUIDE.md → Troubleshooting section

---

Ready to trade! 🚀