16 skills · category

Documents & Files skills for Claude Code

Skills for producing and parsing real office files: PDF extraction, merging and form filling, Word documents that keep their styles, Excel workbooks with working formulas, PowerPoint decks, OCR for scans, true redaction, LaTeX, Mermaid diagrams, and pandoc publishing.

Download all 16 (.zip) Browse the list

Every documents & files skill

Each card opens the full playbook, example prompts and install commands.

No. 366

Contract Analysis

Cited clause tables and deadlines, not legal advice

Claude turns a contract into a cited reading aid: defined terms, a clause table with section numbers and quoted operative words, and an obligations table with computed renewal and notice dates. A clause checklist drives rated risk flags from your side of the deal, missing clauses are listed, and every report opens with a not-legal-advice notice and ends with questions for counsel.

contract-analysis 164 lines · 2 files

★ No. 367

Document Redaction

Remove the data, not just hide it under a box

Black boxes over live text are how redacted court filings leak. With this skill Claude deletes the underlying characters and image pixels using PyMuPDF, scrubs metadata, comments, tracked changes and attachments, and rewrites the file in full. A separate verifier searches two text extractors, raw bytes, fields and metadata, and every job ends with a redaction log.

document-redaction 179 lines · 3 files

No. 368

Word Documents (.docx)

Styled, editable Word files that survive review

Claude builds Word files the way a careful editor would: named styles and real headings instead of bold Normal text, tables and images sized for the page, and templates that keep branding intact. Before editing, a bundled inspector reports tracked changes, comments and fields, so revisions are never silently mangled, and replacements that span split runs keep their formatting.

docx-documents 216 lines · 2 files

No. 369

EPUB Ebooks

Ebooks that pass epubcheck and read well anywhere

Claude structures a manuscript into chapters with semantic markup and alt text, writes metadata and reader-friendly CSS, and builds the EPUB with one pandoc command. A bundled linter checks mimetype, OPF manifest, spine, nav and broken links in seconds before epubcheck and Ace by DAISY run, and the book is opened in real readers before it goes to KDP, Apple Books or Kobo.

epub-ebooks 205 lines · 2 files

No. 370

Financial Models

Live formulas, one input sheet, a balance check at zero

Claude builds models the way a careful analyst does: every input on one labelled sheet, one copied formula per row across a shared monthly time axis, an MRR waterfall with churn applied to opening balances, and linked statements whose balance check reads zero in every period. A bundled linter finds numbers typed inside formulas, constants on calculation sheets and inconsistent rows before anyone sees the file.

financial-models 259 lines · 2 files

No. 371

Google Sheets Automation

Batched, idempotent Sheets jobs that respect quotas

Claude reads and writes Google Sheets in a few batched calls with unformatted values, keeps code-owned ranges separate from columns people edit, and upserts by key instead of row position. Apps Script code works on arrays with one read and one write, uses LockService and continuation triggers to stay under the 6-minute limit, and every job retries 429s with backoff.

google-sheets-automation 187 lines

No. 372

HTML to PDF

Print CSS that breaks pages where you meant

Claude chooses the rendering engine from the job: WeasyPrint for templated business documents, Playwright when JavaScript or browser fidelity matters, Paged.js for book layouts. Templates use @page rules, margin-box counters, named pages and break controls from the start, assets are local so fonts embed, and every PDF is checked with pdffonts and page images.

html-to-pdf 183 lines

No. 373

Invoice Generation

Gapless numbers, exact cents, immutable invoices

Claude models invoices as drafts that become immutable when issued, assigns gapless numbers per series inside the issuing transaction, and computes totals with Decimal under one written rounding policy shared by PDF, XML and ledger. Corrections go through credit notes, e-invoice formats come from the same frozen data, and archived files carry a hash in write-once storage.

invoice-generation 190 lines

No. 374

LaTeX Documents

Clean builds, resolved citations, no ?? in the PDF

Claude sets up LaTeX projects that build with one latexmk command: a sane preamble with packages in the right order, biblatex with biber, booktabs tables, siunitx units, and cleveref cross-references. After each build a bundled log parser lists errors by line, undefined references and citations, and wide overfull boxes, so fixes target causes instead of rerunning and hoping.

latex-documents 211 lines · 2 files

No. 375

Markdown Publishing

One Markdown source, clean PDF, DOCX and HTML

Claude sets up repeatable pandoc builds instead of one-off commands: a defaults file per output, a reference.docx that carries your Word house styles, citeproc with real CSL styles, and the right PDF engine for the job, whether LaTeX, Typst or WeasyPrint. It handles cross-references, math, diagrams and Lua filters, then opens each output to check it.

markdown-publishing 190 lines

No. 376

Mermaid Diagrams

Diagrams as code that render on the first try

Claude picks the diagram type that answers the question, builds it from the real code or schema, and keeps each view under about twenty nodes. It avoids the parser traps that blank a diagram, like a lowercase end label or unquoted parentheses, themes with frontmatter config that survives dark mode, and renders with mmdc before handing it over, accounting for older Mermaid versions on GitHub.

mermaid-diagrams 219 lines · 2 files

No. 377

OCR for Scans

Searchable scans with measured, not assumed, accuracy

Claude OCRs only the pages that need it, preprocesses photos and scans for Tesseract by fixing rotation, contrast, scale and borders, and picks the page segmentation mode that matches the layout. A bundled script reports word-level confidence and the shakiest words, so numbers and names get checked against the image, and tables or handwriting are routed to better tools.

ocr-scans 164 lines · 2 files

No. 378

PDF Toolkit

Inspect first, then extract, merge, fill or generate

Claude inspects every PDF before touching it, checking for a text layer, encryption, form fields and XFA, then picks the right library for the job. Tables come out as clean CSV rows with totals checked, merges keep outlines, forms get filled using their exact field names, and new documents are built with reportlab. Every output is reopened and verified rather than assumed.

pdf-toolkit 210 lines · 2 files

No. 379

PowerPoint Decks (.pptx)

On-brand slides you've actually looked at

Claude writes a takeaway-titled outline first, then builds slides from your template's layouts by name, filling placeholders rather than scattering text boxes. Charts are native and editable, notes carry the detail, and theme fonts stay intact. A bundled linter flags missing titles, overflow and empty placeholders, and every deck is rendered to images and reviewed before handoff.

pptx-decks 194 lines · 2 files

No. 380

Report Automation

Recurring reports that refuse to send bad numbers

Claude turns a hand-built weekly or monthly report into one command keyed on the reporting period: versioned SQL, validation gates for freshness, completeness and reconciliation, deterministic matplotlib charts, and Jinja2 templates rendered to PDF, HTML email or Slack. Runs are idempotent, never resend by accident, print their template version, and leave a manifest for audit.

report-automation 218 lines

No. 381

Excel Spreadsheets (.xlsx)

Live formulas, real number formats, zero #REF!

Claude writes workbooks a finance analyst would trust: real dates and numbers with formats, totals as auditable formulas, frozen headers, tables, dropdown validation and charts. It knows openpyxl never calculates, so it never saves over a data_only load, flags formulas without cached values, and runs a bundled audit for error cells and external links before reporting numbers.

xlsx-spreadsheets 207 lines · 2 files

Choosing Documents & Files skills

Install the skills that match requests you make every week; each costs about 100 tokens of standing context until it loads. If you are not sure where to begin, start with Document Redaction. For background, read Using Skills on claude.ai, the Claude API, and the Agent SDK, The Best Claude Code Skills in 2026, by Use Case and How to Install Claude Code Skills (5 Ways).

By technology: PDF skills Python skills

Install several

Install several at once

The installer accepts any number of skill names. Add --project to install into the current repository.

curl -fsSL https://claudeskills.page/install.sh | sh -s -- contract-analysis document-redaction docx-documents epub-ebooks financial-models google-sheets-automation html-to-pdf invoice-generation latex-documents markdown-publishing mermaid-diagrams ocr-scans pdf-toolkit pptx-decks report-automation xlsx-spreadsheets

Documents & Files skills FAQ

How do I install every Documents & Files skill at once?

Download the bundle at https://claudeskills.page/downloads/category/documents.zip and unzip it into ~/.claude/skills/ (or .claude/skills/ inside a repository). To pick a few, pass several names to the installer: curl -fsSL https://claudeskills.page/install.sh | sh -s -- contract-analysis document-redaction docx-documents.

Will installing 16 skills slow Claude Code down?

No. Claude Code keeps only each skill's short description in context — roughly 100 tokens per skill — and loads the full instructions only when a request matches. Install the whole category if it fits your work; unused skills cost almost nothing.

Which Documents & Files skill should I start with?

Start with Document Redaction. They cover the most common requests in this category; add the rest as your work calls for them.