Hi everyone,


I’m currently working on a project called the YV Strategy Audit Engine.


The basic idea is simple:


Before converting a trading strategy into an MQL5 EA, I want to validate whether the strategy specification is actually complete and executable at the broker/execution level.


The tool is designed to analyze:


• Raw strategy logic• Entry / Exit conditions• SL / TP and position sizing• Indicator calculations• Timeframe and candle timing• Broker specifications• Spread, tick size, tick value and execution requirements• Pine Script ↔ MQL5 equivalence• Missing or ambiguous specifications• Execution-sensitive areas


One important design principle is that the tool should NOT modify or rewrite the trading code.


It should only identify gaps, inconsistencies and potential implementation risks, then provide suggestions.


I’m currently testing the audit engine using real MQL5 EAs and blind validation runs. One of the recent tests showed an interesting problem: a static extractor could detect that something was missing even when the functionality actually existed deeper inside helper functions.


That led me to focus more on data-flow tracing, function scope, deferred SL/TP handling and hand-coded indicators rather than relying only on keyword/API detection.


I’ve attached a visual overview of the project.


I’d be interested in feedback from experienced traders and EA developers:


1. What specifications do you usually define before developing an EA?2. What implementation details have caused unexpected differences between backtests and live trading?3. Which broker/execution parameters do you consider essential?4. Do you think a pre-development strategy audit would be useful in your own workflow?


I’m mainly looking for technical feedback and real-world experience.


Thanks.

Lampiran:

attachedImage
Consistency before scale.