PDF Tools Skill

Extract text, merge, split, and manipulate PDF files.

When to Use

✅ USE this skill when:

• "Extract text from this PDF"
• "Merge these PDFs together"
• "Split this PDF into pages"
• "Add watermark to PDF"
• "Convert PDF to images"

When NOT to Use

❌ DON'T use this skill when:

• Creating PDFs from scratch → use reportlab/fpdf
• Complex PDF editing → use dedicated PDF software
• PDF forms → use specialized tools

Setup

Install required packages:

pip install pypdf pdf2image pillow

For PDF to image conversion, also install poppler:

• macOS: brew install poppler
• Linux: sudo apt install poppler-utils

Commands

Extract Text

{baseDir}/pdf-text.sh document.pdf
{baseDir}/pdf-text.sh document.pdf --pages 1-5
{baseDir}/pdf-text.sh document.pdf --out extracted.txt

Merge PDFs

{baseDir}/pdf-merge.sh file1.pdf file2.pdf --out merged.pdf
{baseDir}/pdf-merge.sh *.pdf --out combined.pdf

Split PDF

{baseDir}/pdf-split.sh document.pdf --out-dir /tmp/pages/
{baseDir}/pdf-split.sh document.pdf --pages 1-3 --out extracted.pdf

Add Watermark

{baseDir}/pdf-watermark.sh document.pdf --text "CONFIDENTIAL" --out watermarked.pdf
{baseDir}/pdf-watermark.sh document.pdf --image logo.png --out watermarked.pdf

PDF to Images

{baseDir}/pdf-to-images.sh document.pdf --out-dir /tmp/pages/
{baseDir}/pdf-to-images.sh document.pdf --pages 1-5 --format png

Get PDF Info

{baseDir}/pdf-info.sh document.pdf
{baseDir}/pdf-info.sh document.pdf --json

Options

• --pages <range>: Page range (e.g., 1-5, 1,3,5)
• --out <path>: Output file path
• --out-dir <dir>: Output directory
• --text <text>: Watermark text
• --image <path>: Watermark image
• --format <fmt>: Image format (png, jpg)
• --json: Output as JSON

Examples

Extract first 10 pages:

{baseDir}/pdf-text.sh report.pdf --pages 1-10 --out summary.txt

Merge all PDFs in directory:

{baseDir}/pdf-merge.sh *.pdf --out combined.pdf

Split PDF into individual pages:

{baseDir}/pdf-split.sh document.pdf --out-dir /tmp/pages/
# Creates: /tmp/pages/page_001.pdf, page_002.pdf, etc.

Add draft watermark:

{baseDir}/pdf-watermark.sh document.pdf --text "DRAFT" --out draft.pdf

Notes

• Text extraction works best with text-based PDFs
• Scanned PDFs require OCR (not included)
• Large PDFs may take time to process
• Watermarks are applied to each page