Parse, structure, and extract content from complex PDF, Word (.docx), Excel (.xlsx/.csv), and PowerPoint files using local terminal tools instead of loading whole binaries into context. Use when the user points at a document file and wants text, tables, metadata, page counts, or bulk extraction across a folder of documents — "pull the tables out of this PDF", "what's in these invoices", "convert this docx to markdown", "read the spreadsheet". Prefer the anthropic-skills pdf/docx/xlsx/pptx skills when the deliverable is a *new or edited* document; use this skill for extraction, triage, and bulk parsing.
