Data Processing Script Generator

Problem addressed

After exporting daily ERP data as CSV or Excel, manual copying, cleaning, mapping, and template reuse can fail because of inconsistent paths, encodings, column names, or formula refresh behavior. This skill targets fixed reporting workflows: it uses an existing Excel template with formatting, slicers, and pivot tables, then turns each day's exported data into a usable report file.

How it works

The workflow is to confirm inputs before generating code. It expects an Excel template and an ERP export sample, then confirms data source, cleaning rules, derived fields, append-or-overwrite behavior, mapping table, field mapping, fixed paths, file naming, and CSV encoding. It then generates a Python script that typically keeps path variables at the top, locates the day's export file by date, tries gb18030, gbk, and utf-8 encodings, loads reference mappings, cleans the data, computes derived fields, and writes into Excel using zip/XML manipulation where possible to preserve formatting, slicers, and pivot tables. Before handoff, it requires a test run and documents where the script lives, how to run it, which path variables to edit, and the expected folder and naming conventions.

Boundaries and cautions

This is best for daily reports with a stable template and repeatable processing rules, not for open-ended analysis where the report structure changes often. If pivot tables are used, enable refresh when the workbook opens. Avoid relying on formulas such as VLOOKUP that do not update automatically when the script writes values; computed values should be written directly. Paths and naming must be confirmed, because unclear conventions make maintenance harder.