---
applyTo: "**/*.py"
---
# Instructions for _Thivyam Python Scripts

- **Purpose:** Python scripts are used for tasks like advanced back-testing, data analysis, machine learning model training, and generating reports from MetaTrader data.
- **Style:** Follow the PEP 8 style guide for all Python code.
- **Libraries:**
    - For data manipulation, use `pandas`.
    - For numerical operations, use `numpy`.
    - For plotting and visualization, use `matplotlib` or `plotly`.
- **Type Hinting:** Use Python type hints for all function signatures to improve code clarity and maintainability.
- **Documentation:** All functions and classes must have clear docstrings explaining their purpose, arguments, and return values.
- **Logging:** Use the standardised `_Thivyam` logging helpers (import from `thivyam.logging`) or Python's `logging` module configured with UTC timestamps for reproducibility.
- **Configuration:** Load credentials and environment-specific settings from `.env` files or OS-level secrets; never hardcode API keys.
- **Testing:** Provide lightweight unit tests (pytest preferred) for data transforms, and include sample datasets when possible.
- **Agent Workflow:** Include a short `README.md` in each script directory explaining inputs, outputs, and any scheduled automation so future agents can run them without guesswork.