Flag Zone Bot is my personal algorithmic trading system for SPY that detects dynamic supply/demand zones and adaptive bull/bear flags across multiple timeframes, executes option orders automatically, and ships real‑time visuals via a Plotly Dash web app. Not financial advice; for educational/personal use only.
- Streams live trades, builds candles (2M/5M/15M …), and maintains the latest price in shared state.
- Calculates EMAs per timeframe and overlays them on live charts.
- Finds zones & levels from 15‑minute history and renders them as bands/lines.
- Detects adaptive flags using slopes (not fixed lines). (See project overview.)
- Executes orders and logs to Discord (order modules not covered here in depth). Overview in README.
- Publishes a web dashboard with:
- Zones Chart (15M history)
- Live 15M, 5M, 2M charts (candles + EMAs)
- Real‑time updates pushed via a FastAPI WebSocket broadcaster.
This system analyzes the market using:
- Dynamic supply & demand zones
- Adaptive bull & bear flags detection via slope algebra, not static lines
- Multiple timeframes for a multi-perspective market view
- Real-time option order execution with automatic scaling in/out
- Discord notifications for each trading step
Note: This project is for personal development only — not financial advice, not for other traders, and not for public use.
✅ Dynamic Zones:
Zones are calculated daily from fresh SPY price data, using an evolving algorithm that continuously improves.
✅ Flag Detection:
Instead of static y=mx+b flag patterns, this version uses adaptable slopes to track price action changes in real time.
✅ EMA Logic:
A 200 EMA filter was tested: trade only flags that align with macro EMA trend.
✅ Fully Automated Orders:
Handles buying options, managing profits with tiered trimming, stop loss logic, and automatic logging to Discord.
✅ Always Improving:
Old versions were heavily overfit. This version’s design is modular — with separate branches for live and experimental logic.
✅ Live Visuals:
Charts for 15-min and 2-min candles update live, showing the bot’s “vision” and calculated zones/levels.
- ✅ Automated Checks:
GitHub Actions runs continuous integration (.github/workflows/python-ci.yml).
- Add more timeframes to expand the bot’s “sightlines.”
- Build a better UI for live status.
- Improve “how it sees” vs. “how it acts” logic split.
- Test more strategies beyond SPY.
┌────────────────────────────────────────────────────────┐
│ Backend │
│ │
Market Data │ ws_auto_connect (Tradier/Polygon) → process_data │
(Trades) ───┼── streams → build candles → write logs (per timeframe) │
│ ↑ latest_price in shared_state │
│ └→ update_ema → update_chart (PNG) │
└────────────────────────────────────────────────────────┘
│ ▲
│ HTTP trigger │ WebSocket push
▼ │
┌────────────────────────────────────────────────────────┐
│ Services │
│ FastAPI (ws_server): │
│ - POST /trigger-chart-update → broadcast "chart:TF" │
│ - WS /ws/chart-updates → clients subscribe │
└────────────────────────────────────────────────────────┘
│
▼
┌────────────────────────────────────────────────────────┐
│ UI (Dash) │
│ Tabs: Zones (15M history), Live 15M/5M/2M charts │
│ On WS message "chart:TF" → regenerates that figure │
└────────────────────────────────────────────────────────┘This project is for educational and personal development only.
It is not financial advice.
No recommendations are made for actual trading.
Flag-Zone-Bot/
├── .git/
├── .github/
│ └── workflows/
│ └── python-ci.yml
├── __pycache__/
├── docs/
│ ├── adr/
│ │ └── 0001-separate-frontend-from-backend.md
│ └── TOC.md
├── indicators/
│ ├── ema_manager.py
│ └── flag_manager.py
├── logs/
│ ├── log files... # SPY_2M.log, SPY_5M.log, SPY_15M.log
│ └── terminal_output.log
├── states/
├── storage/
│ ├── csv/
│ │ ├── order_log.csv
│ │ └── SPY_15_minute_candles.csv
│ ├── data/
│ │ ├── 2m/
│ │ │ └── 2025-09-02.parquet # Alot of Parquet Files
│ │ ├── 5m/
│ │ │ └── 2025-09-02.parquet # Alot of Parquet Files
│ │ └── 15m/
│ │ └── 2025-09-02.parquet # Alot of Parquet Files
│ ├── emas/
│ │ ├── 2M.json
│ │ ├── 5M.json
│ │ └── 15M.json
│ ├── flags/
│ │ ├── 2M.json
│ │ ├── 5M.json
│ │ └── 15M.json
│ ├── images/ # this is where everything chart/image-wise is saved
│ │ ├── SPY_2M_chart.png
│ │ ├── SPY_5M_chart.png
│ │ ├── SPY_15M_chart.png
│ │ └── SPY_15M-zone_chart.png
│ ├── markers/
│ │ ├── 2M.json
│ │ ├── 5M.json
│ │ └── 15M.json
│ ├── objects/
│ │ │ └── 15m/
│ │ ├── objects.json
│ │ └── timeline.json
│ ├── duck.py
│ ├── message_ids.json
│ ├── parquet_writer.py
│ ├── viewport.py
│ ├── week_ecom_calendar.json
│ └── week_performances.json
├── strategies/
│ └── trading_strategy.py
├── tests/ # This is for `Unit Tests`
│ ├── storage_unit_tests/
│ │ ├── conftest.py
│ │ ├── test_compaction.py
│ │ ├── test_csv_to_parquet_days.py
│ │ ├── test_parquet_writer.py
│ │ └── test_viewport.py
│ └── purpose.md
├── tools/
│ ├── __init__.py
│ ├── compact_parquet.py
│ ├── csv_to_parquet_days.py
│ ├── generate_structure.py
│ └── plot_candles.py
├── utils/
│ ├── __pycache__/
│ ├── data_utils.py
│ ├── ema_utils.py
│ ├── file_utils.py
│ ├── json_utils.py
│ ├── log_utils.py
│ ├── order_utils.py
│ └── time_utils.py
├── venv/ # You need to make this then `pip install -r requirements.txt`
├── web_dash/
│ ├── __init__.py
│ ├── dash_app.py
│ ├── chart_updater.py
│ ├── ws_server.py
│ ├── about_this_dash_folder.txt
│ ├── charts/
│ │ ├── live_chart.py
│ │ └── zones_chart.py
│ └── assets/
├── .gitignore
├── buy_option.py
├── config.json
├── cred.py
├── data_acquisition.py
├── economic_calender_scraper.py
├── error_handler.py
├── main.py
├── objects.py
├── order_handler.py
├── paths.py
├── print_discord_messages.py
├── README.md # THIS
├── requirements.txt
├── rule_manager.py
├── sentiment_engine.py
├── shared_state.py
└── submit_order.pyBuilt and maintained by [Merced3]
Always learning. Always building.