Document Scanning

This handles the grunt work of finding Office documents and PDFs across your project, filtering out the temp files and lock files that clutter up scan results, and extracting metadata for accessibility audits. You get PowerShell and Bash commands for recursive and non-recursive discovery, git-based delta detection to see what changed since your last commit or tag, and a clean context format for passing files to specialized sub-agents for Word, Excel, PowerPoint, and PDF checks. It's built around three scan profiles (strict, moderate, minimal) that let you tune severity filters depending on whether you're triaging a massive document library or preparing public-facing content. The exclusion patterns alone save you from chasing down a dozen tilde-prefixed noise files every time you run a scan.

Install to Claude Code

npx -y skills add community-access/accessibility-agents --skill document-scanning --agent claude-code