Word Report Auto-Formatting

The Problem It Solves

Word documents generated by AI or as initial drafts often suffer from inconsistent formatting and punctuation errors, undermining their professionalism. The most insidious issue affecting deliverable quality is the confusion between full-width and half-width Chinese punctuation—using visually similar half-width punctuation within Chinese text, which violates the national standard GB/T 15834—2011. Furthermore, manually adjusting page setup, fonts, line spacing, and heading hierarchies to meet the strict formatting standards of institutions like Bowei Consulting is a tedious and error-prone task.

Core Capabilities and Workflow

This skill translates formatting standards into executable, automated rules. It is built upon the Bowei Consulting Word Report Formatting Standard V2.6 and the national standard, using Python scripts to batch-correct and audit documents.

Key operational modes include:
- Auto-Formatting: Reads the document and automatically adjusts page layout, fonts, line spacing, and heading styles according to the standard, enabling Chinese punctuation correction by default. The format_word.py script safely converts the majority of half-width Chinese punctuation to full-width.
- Formatting Audit: Runs audit_word.py to perform a comprehensive check, generating a detailed issue list including punctuation errors P1–P20.
- New Report Creation: Ensures source format compliance when building a document from scratch.

The Intelligent Logic for Punctuation Correction:
1. Safe Automatic Correction: E.g., converting half-width commas, periods, brackets, and quotation marks within Chinese text to full-width.
2. Contextual Rule: If any character adjacent to a punctuation mark is Chinese, that punctuation must be full-width.
3. Whitelist Preservation: Half-width punctuation within URLs, file paths, version numbers, chemical formulas, etc., is not incorrectly altered.

The script first handles absolutely safe corrections. For ambiguous cases, such as whether a mid-sentence half-period is part of a version number or if a single em dash is a legal separator, it flags them in the audit report for manual human judgment, avoiding false positives.

Scope and Important Considerations

This is a formatting tool for a specific standard, not a universal formatter:
- Standard-Locked: The current rules are strictly tied to Bowei Consulting's V2.6 standard. To adapt to other corporate or government standards (e.g., GB/T 9704-2012 for Party/Government Official Documents), you must modify the constant values within the scripts.
- Requires Human Collaboration:
- Cover Page Layout: Complex cover designs may require manual fine-tuning.
- Table of Contents Page Numbers: After the script inserts the TOC, you must right-click it in Word and “Update Field” -> “Update entire table” to get accurate page numbers.
- Certain Punctuation: Items flagged as INFO or requiring judgment in the audit report (e.g., a single em dash) need semantic analysis.
- Table Headers on Page Breaks: Due to library limitations, repeating the header on the second page of a table that spans pages must be set manually.
- Font Dependencies: Fonts like FangSong, KaiTi, SimHei, and FangZheng XiaoBiaoSong must be correctly installed in the system environment where the script runs.
- Coordination with stop-slop-zh: This skill addresses AI-generated artifacts at the format and punctuation level. To simultaneously remove AI-style sentence structure and wording, it is recommended to process the text with the stop-slop-zh skill first, then apply this skill for formatting.