ZONES

MT4 trading system with a Python dashboard, named-pipe DLL bridge, and customer-specific SMC / ICT market structure logic.

image source

What ZONES Is

ZONES is a hybrid MT4 + Python trading application built around customer-defined market structure rules.

Current foundation in this repo includes:

• ZigZag-driven Temp Zones
• Fractal-promoted Main Zones
• S1 / S2 / S3 main-zone strength classes
• HH / LH / LL / HL / EQHH / EQLL structure labeling
• BOS, CHOC, and RRR event flow
• candle-body buy/sell mode logic
• support, resistance, and liquidity mapping
• multi-timeframe analysis for 1H, 5M, and 1M
• browser dashboard with structured route views
• browser candle terminal with ZONES overlays filtered by selected symbol and timeframe
• TradingView Advanced Chart page with live symbol switching, public drawing tools, and ZONES projection overlays
• browser backtesting for zone-touch entries with risk, drawdown, R-multiple, and trade-list metrics
• hedge fund style portfolio analysis with exposure, leverage, concentration, VaR proxy, and PnL attribution metrics
• system status page for runtime settings, remote DB control, AI training, and feedback capture
• editable allowed-session control on the system page for the execution gate
• local MT4 bridge using a localhost WebSocket as the primary transport, with the named-pipe path kept for legacy compatibility
• local SQLite by default, with optional remote PostgreSQL support

Architecture

The repo currently uses this flow:

1. MT4 EA collects candles, account data, and symbol context.
2. ZONES.mq4 sends requests through the local WebSocket bridge, while the named-pipe server remains available for legacy clients.
3. zones.py runs the Python service and dashboard.
4. supply_demand_ea analyzes structure, zones, execution filters, and phase outputs.
5. The dashboard renders readable HTML views for live monitoring.
6. The MT4 chart fetches the latest analyzed overlay from Python, so chart zones and backend decisions come from the same ZigZag / Fractal report.

Key files:

• Python entrypoint: zones.py
• Engine: supply_demand_ea/analysis.py
• Browser dashboard: supply_demand_ea/dashboard.py
• Live bridge service: supply_demand_ea/bridge.py
• Pipe server: supply_demand_ea/ipc.py
• MT4 EA: MQL4/Experts/Zones/Zones.mq4
• MT4 include: MQL4/Include/ZonesBridge.mqh
• DLL source: bridge_dll/ZonesBridgeDll.cpp
• Prebuilt x86 debug DLL: bridge_dll/bin/x86/Debug/ZonesBridge.dll
• PDF rulebook summary: docs/zones_pdf_rulebook.md

Customer Rulebook Status

The Python engine has been refactored around the provided PDF requirements and now models:

• temp-zone creation from ZigZag structure points
• main-zone promotion from Fractal confirmation
• discrete main-zone strengths instead of generic scores only
• body-based zone mode detection
• CHOC deletion of unconfirmed temp zones
• RRR stages: respect, reject, retest, close_confirm
• structural invalidation language and structural trailing references
• MT4 chart rendering from the same analyzed ZigZag / Fractal overlay that powers the browser dashboard

Supporting docs:

• docs/mt4-python-architecture.md
• docs/zones_pdf_rulebook.md
• docs/supply_demand_ea_phase1.md
• docs/live_mt4_bridge.md
• docs/mt4_dll_bridge.md
• docs/chart_terminal.md
• docs/backtesting.md
• docs/runtime_settings.md
• docs/installation_package.md
• docs/INSTALL.md
• docs/RELEASE.md
• docs/TROUBLESHOOTING.md
• docs/GITHUB_RELEASE_CHECKLIST.md

Run The App

Start the Python service:

python zones.py

Open the browser dashboard:

http://127.0.0.1:8787

Useful routes:

• /
• /chart
• /backtest
• /portfolio
• /system
• POST /api/ingest
• /api/analysis
• /api/reports
• /api/symbols
• /api/commands
• /api/backtest
• /api/portfolio
• /api/health
• /api/schema
• /api/system/status
• /api/chart

Browser routes render structured HTML by default.

For raw JSON, add:

?format=json

Live MT4 or bridge clients can post JSON snapshots directly to:

http://127.0.0.1:8787/api/ingest

Browser Chart And MT4 Commands

Open the chart page:

http://127.0.0.1:8787/chart

What it includes:

• local ZONES candle terminal rendered from the app payload
• supply, demand, support, and resistance overlays drawn directly on the candles
• candlestick chart for live or waiting candle data
• TradingView Advanced Chart widget as the main browser chart surface
• drawn zone overlays from the ZONES engine
• tracked-symbol and custom-symbol switching on the chart page
• TradingView symbol override input for provider-specific symbols like FX:EURUSD or OANDA:XAUUSD
• timeframe switching for available feeds
• pending command queue view
• execution history view
• browser form that queues MT4 commands through the existing DLL bridge flow
• auto-filled entry, SL, and TP from the current setup, with manual adjustment before queueing
• basic validation so empty lot, price, ticket, or alert message inputs are rejected before hitting MT4

Symbol workflow:

1. Start python zones.py.
2. Open http://127.0.0.1:8787/chart.
3. Use Tracked Symbol to load a symbol already ingested from MT4, or type a custom symbol such as EURUSD.
4. Pick the timeframe tab shown above the terminal.
5. Review the local ZONES candle terminal. The rectangular overlays are the exact zones from the selected symbol payload.

The browser command panel can queue:

• market_buy
• market_sell
• buy_limit
• sell_limit
• buy_stop
• sell_stop
• modify_ticket
• close_ticket
• delete_ticket
• close_all
• alert

Those commands are picked up by the MT4 EA when it polls the local bridge.

Backtesting

Open:

http://127.0.0.1:8787/backtest

The backtester uses the selected symbol's chart_data and filtered ZONES overlays. It simulates a zone-touch strategy where the signal candle touches an eligible zone and the entry happens at the next candle open.

Metrics include:

• ending balance, net PnL, and return percentage
• win rate, profit factor, expectancy in R, and net R
• max drawdown percentage
• full trade list with entry, stop, target, exit reason, R multiple, and PnL

Raw JSON is available at:

http://127.0.0.1:8787/api/backtest?symbol=EURUSD&timeframe=5M&format=json

Portfolio Analysis

Open:

http://127.0.0.1:8787/portfolio

The portfolio page shows hedge fund style metrics from the latest account snapshot:

• NAV and daily return proxy
• gross and net exposure in lots
• long/short lots and long/short ratio
• leverage proxy
• margin level and margin utilization
• concentration risk, HHI concentration, and effective bets
• 95% VaR proxy and 2% stress-loss proxy
• PnL attribution by long and short books
• exposure by symbol

Raw JSON is available at:

http://127.0.0.1:8787/api/portfolio?format=json

MT4 Setup

For the current local bridge, MT4 uses DLL imports instead of WebRequest.

1. Copy bridge_dll/bin/x86/Debug/ZonesBridge.dll into your MT4 MQL4/Libraries folder as ZonesBridge.dll.
2. Copy or compile MQL4/Experts/Zones/Zones.mq4 in MetaEditor.
3. Ensure MQL4/Include/ZonesBridge.mqh is available.
4. In MT4, enable Allow DLL imports.
5. Start python zones.py.
6. Attach the ZONES EA to a chart.

Notes:

• MT4 must use the 32-bit DLL build.
• The Python app should be running before the EA starts posting data.
• The dashboard can ingest all symbols from Market Watch while drawing on the attached chart.
• When PostAllMarketWatchSymbols is enabled, the EA now polls command queues across Market Watch symbols too, so browser orders are not limited to the attached chart symbol.
• The attached MT4 chart now draws zones, swings, and structure markers from the Python analysis snapshot instead of a second local swing algorithm.
• If you update the DLL source, rebuild the x86 DLL and replace MQL4/Libraries/ZonesBridge.dll before testing in MT4.

Build The DLL

The DLL project is in bridge_dll.

Open that folder in VS Code or Visual Studio and build the x86 target.

Helpful files:

• bridge_dll/CMakeLists.txt
• bridge_dll/CMakePresets.json
• bridge_dll/build_msvc.bat
• bridge_dll/build_mingw.bat
• bridge_dll/README.md

Expected output:

bridge_dll/bin/x86/Debug/ZonesBridge.dll

Database

Default database:

SQLite

Default file:

logs/zones.sqlite

Optional remote database:

$env:ZONES_DATABASE_URL = "postgresql://user:password@host:5432/zones"
python zones.py

SQLite URLs also work:

$env:ZONES_DATABASE_URL = "sqlite:///C:/path/to/zones.db"

The browser system page can also switch the runtime repository without editing environment variables:

• open /system
• paste a SQLite or PostgreSQL URL
• enter the database username and password separately when the URL requires credentials
• save runtime settings

The same /system page also stores Telegram settings:

• paste the Telegram bot token
• send one message to the bot from the Telegram account or group you want to use
• save runtime settings, and ZONES will call Telegram getUpdates to discover the chat id automatically
• confirm the discovered chat id in the system status card

For chart viewing, /system can store a default chart symbol, and /chart also lets the user select a tracked symbol or type a custom symbol before viewing the ZONES candle terminal.

The same page also exposes live zone sizing controls:

• Temp Zone Min Thickness
• Temp Zone Max Thickness
• Main Zone Min Thickness
• Main Zone Max Thickness

AI Training And Signal

ZONES now includes a lightweight trainable signal model for runtime status and signal guidance.

What it does today:

• trains from recorded trade feedback when enough samples exist
• falls back to bootstrap training from stored reports and execution decisions
• scores the latest live report with a current AI signal and confidence
• exposes model state on /system and /api/system/status

To use it:

1. ingest live MT4 reports or let the app collect reports normally
2. open /system
3. record feedback rows for wins, losses, and PnL
4. click Train Signal Model

Tests

Run the test suite:

python -m unittest discover -s tests -v

The current tests cover:

• multi-timeframe analysis
• live ingest contracts
• named-pipe health and ingest processing
• command queue flow
• browser HTML route rendering
• temp-to-main promotion
• RRR stage generation
• candle-body mode logic
• CHOC deletion behavior

Repo Notes

A few legacy files and folders still exist in the repo from earlier project history, but the current active application path is:

• zones.py
• supply_demand_ea/
• MQL4/Experts/Zones/
• MQL4/Include/
• bridge_dll/
• docs/
• tests/

License

See LICENSE and LICENSE.md.