python-workspace

Translate MQL5 indicators to Python by exporting reference data from MT5, implementing the indicator logic in Python, then validating the output against the original using correlation and error metrics. This Skill automates the validation workflow with a 32-test suite and DuckDB regression tracking to ensure functional equivalence.

Yes, validate Python implementations against MQL5 references using autonomous validation that checks correlation ≥0.999, MAE <0.001, and zero NaN counts after a 5000-bar warmup. The Skill applies structured testing across headless exports and GUI-based indicator workflows.

Accuracy requirements include correlation ≥0.999 between Python and MQL5 outputs, mean absolute error <0.001, and zero NaN counts after warmup. The Skill enforces these thresholds across a 32-test validation suite and DuckDB-driven regression tracking.

Export MQL5 data using headless data export or GUI-based indicator export workflows within MT5, then feed the reference output into the Skill's Python validation pipeline. The validation loop confirms functional equivalence through structured test coverage and regression analysis.

Yes, the Skill handles end-to-end migration workflows including Laguerre RSI, ATR, and custom validation pipelines. It applies the same 5000-bar warmup, correlation, and error thresholds across all indicator types.

The Skill uses a self-correcting loop with clear boundaries to identify divergence. It tracks mismatches through DuckDB regression analysis and reports failures against the 32-test validation suite, pinpointing where Python logic deviates from MQL5 behavior.