Back to Skills

PDF Tools

Read, extract, merge, split, rotate, watermark, encrypt, and OCR PDF files — including filling PDF forms and extracting tables

pdfdocumentsofficialanthropicfiles

Skill Content

# PDF Tools Skill

Official Anthropic skill for working with PDF files. Triggers whenever a .pdf file is mentioned as input or output.

## Capabilities

- **Read and extract** text and tables from PDFs
- **Merge** multiple PDFs into one
- **Split** a PDF into per-page or per-section files
- **Rotate** individual pages
- **Watermark** with text or image overlays
- **Encrypt / decrypt** with passwords
- **Fill PDF forms** programmatically
- **Extract images** from embedded resources
- **OCR** scanned PDFs so their content becomes searchable

## When It Triggers

- User mentions a .pdf file by name
- User asks to produce a PDF report or deliverable
- User asks to combine, split, or manipulate PDFs

## Installation

Included in the `anthropic-skills` plugin:

```bash
/plugin install pdf@anthropic-skills
```

## Example Prompts

- "Merge these three PDFs into one"
- "Extract the table from page 4 of report.pdf"
- "OCR this scanned invoice"
- "Fill out the W-9 form with my info"
- "Add a watermark to every page"

## Implementation

Uses pypdf, pdfplumber, and pytesseract under the hood. No external services — runs locally.

How to use

  1. Copy the skill content above
  2. Create a .claude/skills directory in your project
  3. Save as .claude/skills/pdf.md
  4. Use /pdf in Claude Code to invoke this skill
View source on GitHub