LVN-Sentinel

LVN Sentinel is a small trading workflow that downloads OHLCV data, scans for LVN-based signals, updates a CSV portfolio, and sends Telegram alerts.

Setup

1. Copy .env.example to .env.
2. Fill in TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID.
3. Install the Python dependencies used by the project:
   • python-dotenv
   • pandas
   • numpy
   • scipy
   • yfinance
   • requests
   • matplotlib (optional: enables the 5-day trend chart image in Telegram; without it a text trend is used)

DATA_DIR, DATABASE_DIR, and LOG_DIR are created automatically on first run. If LOG_FILE is left empty, the app writes logs to LOG_DIR/app.log.

Usage

Run the sentinel from the repository root:

python main.py

Daily P&L report

Each run sends the Telegram report LVN Sentinel Report - rev 031026 with daily realized P&L, unrealized P&L of open positions, Total P&L from first day run and a 5-day trend (text, plus a PNG chart when matplotlib is installed).

• Realized P&L comes from DATABASE_DIR/history.csv (closed events).
• Unrealized P&L is computed from open positions in DATABASE_DIR/portfolio.csv; if today's price is missing, the last known price (last_price) is used and the ticker is flagged in the report.
• A daily snapshot is stored in DATABASE_DIR/daily_pnl.csv (one row per day; repeated runs on the same day overwrite that day's row, so nothing is double-counted).