PERMACT / MT5 VALIDATION LAB

Build it. Break it. Verify it.

Reproducible MetaTrader 5 and MQL5 tutorials for people who want to understand what an Expert Advisor actually does before they trust a backtest—or risk real capital.

9
reproducible labs
1
versioned starter EA
0
profit promises

START HERE / FOUR PASSES

One workflow, from blank file to skeptical review

LAB NOTES

Latest MT5 validation labs

System Trading

Using AI to Evaluate an MT5 Strategy: Bollinger Bands, RSI, and Backtest Iteration

Turn a Bollinger Bands and RSI idea into an AI-ready MT5 specification, analyze backtest evidence, and use result-specific prompts to compare strategy versions.

Published 18 min read
System Trading

Build a Testable MT5 EA with AI: MQL5 Specs and Prompts for Codex or Claude

Turn an MT5 strategy idea into a testable MQL5 specification for Codex or Claude Code, with risk limits, recovery rules, evidence, and validation gates.

Published 15 min read
System Trading

How to Recover MT5 EA State After Restart or Reconnect

Design a restart-safe MT5 EA that reconciles positions, pending orders, and trade history before it can send another signal after reconnect.

Published 18 min read
System Trading

How to Forward-Test an MT5 EA on a Demo Account

A reproducible checklist for forward-testing an MT5 Expert Advisor on demo: uptime, duplicate orders, protective stops, execution drift, and stop criteria.

Published 16 min read
System Trading

Build and Run Your First MQL5 Expert Advisor on an MT5 Demo Account

Download, compile, attach, and inspect a small MQL5 Expert Advisor in MetaTrader 5 while learning permissions, inputs, logs, and demo-only safety limits.

Published 18 min read
System Trading

How to Read an MT5 Strategy Tester Report

Audit an MT5 backtest report from Journal to drawdown, Profit Factor, trade history, and equity curve—without relying on a single headline metric.

Published 16 min read

CODE & DOWNLOADS

Read the source before you run the robot

The Permact First Candle EA is deliberately small. It exposes lot size, stop distance, take-profit distance, buffer, and magic number so each test condition can be recorded and challenged.

CHECK BEFORE RUN

[ ] source inspected
[ ] compile: 0 errors
[ ] symbol rules recorded
[ ] Journal checked
[ ] unseen period tested
[ ] demo behavior reviewed
Evidence before confidence.

Every English lab is edited as technical education: observable steps, official references, explicit assumptions, and correction paths. It is never a broker endorsement or a claim that an EA will be profitable.

Read our testing and editorial method