Public CORE repository for study, analysis, and demonstration.
This project started as a practical, hands-on effort: reviewing and testing many “robots” and approaches to understand what is real vs. noise in live markets.
Over time, the production system evolved into something much larger than a public demo:
- It accumulated operational tooling, validation workflows, and safety constraints.
- It incorporated research and implementation details that are not appropriate to publish in full.
- It became “operationally sensitive” (easy to misuse or misrepresent if copied without context).
That is why the project is now split:
- CORE (this repo): a minimal, auditable demo for learning.
- PREMIUM (private): the production codebase and operational stack.
If you’re here to learn, CORE is intentionally kept small and readable.
- A simplified MQL5 demo EA intended for educational purposes.
- A public snapshot that intentionally excludes proprietary production components.
- Not a production trading system.
- Not allowed for live/funded trading.
- Not allowed for broker-connected paper trading or any setup that mirrors live execution.
See:
TRADING_RESTRICTIONS.mdDATA_SOURCES.md
Start with these files (in order):
TRADING_RESTRICTIONS.md— non-negotiable usage restrictions.DATA_SOURCES.md— dataset policy (no redistribution).MQL5/Experts/AurumCore.mq5— the entire demo EA source.MQL5/Indicators/— educational indicators (spread, session windows, volatility).docs/PREMIUM_NOTE.md— sanitized overview of what exists privately.
- Open
MQL5/Experts/AurumCore.mq5in MetaEditor. - Compile.
- Attach the EA to a chart.
- Keep
InpAllowTrading=falseby default.- If you want to observe order flow mechanics in a local sandbox, you can toggle it to
true.
- If you want to observe order flow mechanics in a local sandbox, you can toggle it to
Safety notes:
- This EA closes its own positions on remove/close (see
OnDeinit). - Uses
MagicNumberto avoid touching other trades.
AurumCore.mq5 intentionally keeps the logic minimal:
- Spread filter: blocks entries when spread exceeds
InpMaxSpreadPoints. - New-bar execution: evaluates signals only on a new bar for
InpTf. - Simple signal: EMA cross (
InpFastMaPeriodvsInpSlowMaPeriod). - Single-position rule: at most one position per symbol+magic.
This is meant to be understandable and auditable in one sitting.
CORE also ships small indicators that help visualize execution constraints and market conditions (without exposing any proprietary strategy logic):
MQL5/Indicators/AurumSpreadMonitor.mq5— plots current spread (points) + a max threshold.MQL5/Indicators/AurumSessionWindow.mq5— draws a daily time window with vertical lines (server time).MQL5/Indicators/AurumVolatilityATR.mq5— plots ATR as a volatility proxy.
Legend:
- ✅ = included
- ❌ = not included
⚠️ = partial / simplified demo
| Area | CORE (public demo) | PREMIUM (private production) |
|---|---|---|
| MQL5 EA source | ✅ | ✅ |
| Educational indicators (spread/session/ATR) | ✅ | ✅ |
| Minimal demo strategy (EMA cross) | ✅ | ❌ |
| Multi-strategy routing/selection | ❌ | ✅ |
| Advanced confluence / proprietary logic | ❌ | ✅ |
| News-aware gating / calendar | ❌ | ✅ |
| Risk management (prop-firm constraints, DD/HWM) | ❌ | ✅ |
| Time-gates / session rules | ❌ | ✅ |
| Slippage/commission/latency modeling | ❌ | ✅ |
| Backtesting engine + runners | ❌ | ✅ |
| Parameter search / optimization | ❌ | ✅ |
| Walk-forward analysis (WFA) | ❌ | ✅ |
| Stress testing / robustness checks | ❌ | ✅ |
| Overfit controls / validation tooling | ❌ | ✅ |
| ML experiments/models (optional) | ❌ | ✅ |
| Datasets redistributed | ❌ | ❌ |
| Live trading allowed | ❌ | ✅ (private only, with ops controls) |
Notes:
- CORE is intentionally constrained so the public code stays reviewable and does not ship operational “go-live” capability.
- PREMIUM details, datasets, and operational infra are intentionally private.
This repo is intentionally narrow-scope. Contributions that fit well:
- New educational indicators (visualization, overlays, regime/volatility tools).
- Bugfixes and clarity improvements for CORE files.
Contributions that will not be accepted:
- Anything that enables broker-connected paper/live trading.
- Anything that attempts to bypass
TRADING_RESTRICTIONS.md. - Proprietary strategy logic from the PREMIUM system.
If you want to request an indicator (MQL5): open an issue with:
- Instrument (e.g., XAUUSD), timeframe, and example screenshots.
- What you want to visualize (not “the strategy”), plus acceptance criteria.
- Telegram (preferred): check my Telegram in my GitHub profile / contact details.
- Repo requests: please open a GitHub Issue using the templates (recommended so requirements stay written).
- Private / commercial: for custom development, partnerships, or investor conversations, reach out on Telegram.
Notes:
- I don’t provide real-time support; async is best.
- No promises of profitability; discussions focus on engineering, risk, and testability.
From time to time, I also take on custom development (paid/private) such as:
- MQL5 indicators tailored to your workflow
- Automation of a manual strategy you already trade, based on your written rules
Important:
- Requests must be well-specified (inputs, outputs, constraints, test cases).
- Anything involving broker-connected execution or live trading is handled privately and is out of scope for this public CORE repository.
- No guarantees of profitability; engineering work focuses on correctness, safety, and testability.
To start: open an issue describing what you want built and what “done” means (recommended). If it’s private/commercial, you can also reach out on Telegram.
- EA (Expert Advisor): automated trading program for MetaTrader.
- CORE: public demo/study repository with minimal logic.
- PREMIUM: private production repository (full system + ops).
- MagicNumber: identifier used to tag orders/positions belonging to this EA.
- Spread: difference between bid and ask (cost); CORE can block entries above a limit.
- Slippage: difference between expected and actual fill price due to execution conditions.
- Commission: per-trade fee model (broker/venue dependent).
- Latency: execution delay that impacts fills and performance.
- Tick vs Bar: tick = every price update; bar = aggregated candle over a timeframe.
- Timeframe (TF): candle period (e.g., M5).
- DD (Drawdown): peak-to-trough equity decline.
- HWM (High-Water Mark): highest equity point used by some trailing DD rules.
- WFA (Walk-Forward Analysis): validate performance across rolling in-sample/out-of-sample windows.
- Overfitting: strategy fits historical noise and fails out-of-sample.
- Prop firm constraints: risk rules imposed by evaluation/funded accounts.
- News gating: reducing or blocking trading around scheduled high-impact news.
MQL5/Experts/AurumCore.mq5— demo EA sourceMQL5/Indicators/— educational indicators
- Python/Nautilus layer is kept private (PREMIUM).
- No market data is redistributed in this repository.
- Telegram: @francomascareloai