allocation-engine

Summary

Version - 0.1.0 - This is hobbyist client side agentic trading package (builds with the allocation-manager service).

1. Install dependencies:

pip install -r requirements.txt

2. Configure credentials:

python scripts/setup_credentials.py

3. Run the trading system (dry run):

python -m trading_system.main

4. Go live when ready:

python -m trading_system.main --live

Structure

robinhood-trading/
├── trading_system/       
│   ├── main.py           # 30-day breakout strategy
│   ├── data_providers/   # Market data (Twelve Data API)
│   ├── strategies/       # Trading strategies
│   ├── state/            # State management
│   └── utils/            # Metrics calculations
│
├── utils/                 # Core utilities
│   ├── rh_auth.py        # Robinhood authentication
│   └── safe_cash_bot.py  # Cash trading bot
│
├── old_strategies/        # Legacy strategies (reference)
├── scripts/               # Utility scripts
└── docs/                  # Additional documentation

For detailed structure, see PROJECT_STRUCTURE.md

30-Day Breakout Strategy:

• Buy when price hits 30-day low
• Sell when price hits 30-day high
• Tracks: BTC, S&P 500 (SPY), QQQ, AMZN
• Persistent state management
• Market data from Twelve Data API

Usage:

# Single run (dry run)
python -m trading_system.main

# Live trading
python -m trading_system.main --live

# Continuous mode (every 5 minutes)
python -m trading_system.main --continuous

# Custom interval (every 10 minutes)
python -m trading_system.main --continuous --interval 10

Full Guide: TRADING_SYSTEM_GUIDE.md

Utility Scripts

# Setup credentials
python scripts/setup_credentials.py

# List accounts
python scripts/list_accounts.py

# Place single order
python scripts/place_single_order.py

# Verify account isolation
python scripts/verify_isolation.py

Legacy Strategies

# Momentum strategy (moving averages)
python old_strategies/momentum_strategy.py

# After-hours strategy
python old_strategies/afterhours_daily_strategy.py

Configuration - on personal RH

Environment Variables (.env)

Before Running -

• Never commit .env to git
• Enable 2FA on your Robinhood account
• Use chmod 600 .env for restrictive permissions
• Dry run mode by default
• Account isolation (locked to 490706777)
• Cash-only trading (no margin)
• 10-second countdown before live trading

# Robinhood Credentials
RH_AUTO_EMAIL=your_email@example.com
RH_AUTO_PASSWORD=your_password
RH_AUTOMATED_ACCOUNT_NUMBER=490706777

# Twelve Data API (for trading_system)
TWELVE_DATA_API_KEY=your_api_key_here

State Management:

• Persistent state in trading_state.json
• Tracks all metrics per symbol
• Records queued and active orders
• Maintains order history

Component Scripts

This is a mess right now, working on clean-up

# Test Twelve Data API
python trading_system/data_providers/twelve_data.py

# Test metrics calculation
python trading_system/utils/metrics.py

# Test strategy logic
python trading_system/strategies/breakout_strategy.py

# Test state management
python trading_system/state/state_manager.py

Core Utility Scripts

from utils.rh_auth import RobinhoodAuth
from utils.safe_cash_bot import SafeCashBot
import robin_stocks.robinhood as r

# Login
auth = RobinhoodAuth()
auth.login('automated')

# Use safe cash bot
bot = SafeCashBot()
bot.get_portfolio_summary()

# Logout
auth.logout()

MISC

• TRADING_SYSTEM_GUIDE.md m
• PROJECT_STRUCTURE.md
• docs/ - Additional documentation

Requirements

• Python 3.7+
• robin-stocks >= 3.0.0
• python-dotenv >= 1.0.0
• pyotp >= 2.9.0
• requests >= 2.31.0

Install all: pip install -r requirements.txt

Workflow

1. Setup (one-time):

pip install -r requirements.txt
python scripts/setup_credentials.py

2. Test in dry run:

python -m trading_system.main

3. Review state:

cat trading_state.json

4. Go live:

python -m trading_system.main --live

5. Run continuously:

python -m trading_system.main --live --continuous

Limits

API Rate Limits

Twelve Data Free Tier:

• 8 calls/minute
• 800 calls/day
• System uses ~12 calls per run (4 symbols × 3 calls)

Usage Limits:

• Continuous mode: 5+ minute intervals
• Daily limit: ~66 runs

Version: 2.0 (Modular Architecture) Last Updated: January 2025