Office Scanning

Office Document Scanning

The scan_office_document MCP tool scans DOCX, XLSX, and PPTX files by parsing their ZIP/XML structure. No external dependencies required.

Rules by Format

Format	Rules	Key Checks
DOCX	16 rules (DOCX-E*, DOCX-W*, DOCX-T*)	Alt text, headings, table headers, language, document title, color-only formatting, empty paragraphs, font sizes
XLSX	14 rules (XLSX-E*, XLSX-W*, XLSX-T*)	Sheet names, merged cells, header rows, chart alt text, defined names, hidden sheets, input messages
PPTX	16 rules (PPTX-E*, PPTX-W*, PPTX-T*)	Slide titles, reading order, alt text, table headers, audio/video, language, font sizes, speaker notes

Usage

# Claude Code
/word-accessibility scan docs/report.docx
/excel-accessibility check data/budget.xlsx
/powerpoint-accessibility review slides/deck.pptx

# Copilot
@word-accessibility scan the annual report
@excel-accessibility check the spreadsheet template
@powerpoint-accessibility review the training presentation

Output Formats

• SARIF (default) - Machine-readable, compatible with GitHub Code Scanning
• Markdown - Human-readable report with severity, rule explanations, and remediation guidance

CI/CD Script

The CI scanner at .github/scripts/office-a11y-scan.mjs:

• Discovers documents recursively (skipping node_modules, .git, vendor)
• Applies .a11y-office-config.json if present
• Outputs SARIF 2.1.0 reports
• Emits GitHub Actions ::error:: and ::warning:: annotations
• Exits with code 1 if error-severity findings are detected

- name: Scan Office documents for accessibility
  run: node .github/scripts/office-a11y-scan.mjs