PDF Processing Skill

When working with PDF files, follow these guidelines:

1. Reading & Extracting from PDFs

For text extraction:

# Extract all text
pdftotext input.pdf output.txt

# Extract specific pages
pdftotext -f 1 -l 10 input.pdf output.txt

# Preserve layout
pdftotext -layout input.pdf output.txt

For extracting images:

# Extract all images
pdfimages -all input.pdf output_prefix

# Extract as PNG
pdfimages -png input.pdf images/page

For metadata:

# Get PDF info
pdfinfo document.pdf

# Get detailed metadata
exiftool document.pdf

2. Creating PDFs

From text/markdown:

# From markdown using pandoc
pandoc input.md -o output.pdf

# From text with formatting
enscript input.txt -o - | ps2pdf - output.pdf

From HTML:

# Using wkhtmltopdf
wkhtmltopdf input.html output.pdf

# With options
wkhtmltopdf --page-size A4 --margin-top 10mm input.html output.pdf

From images:

# Convert images to PDF
convert image1.png image2.png output.pdf

# Multiple images
img2pdf img1.jpg img2.jpg -o output.pdf

3. Merging PDFs

# Merge multiple PDFs (using pdftk)
pdftk file1.pdf file2.pdf file3.pdf cat output merged.pdf

# Using ghostscript
gs -dBATCH -dNOPAUSE -q -sDEVICE=pdfwrite -sOutputFile=merged.pdf file1.pdf file2.pdf

# Using pdfunite
pdfunite file1.pdf file2.pdf output.pdf

4. Splitting PDFs

# Split into individual pages
pdftk input.pdf burst output page_%02d.pdf

# Extract specific pages
pdftk input.pdf cat 1-5 output first-5-pages.pdf

# Extract page ranges
pdftk input.pdf cat 1-10 25-30 output selected.pdf

5. Converting PDFs

PDF to Images:

# To PNG (high quality)
pdftoppm -png -r 300 input.pdf output

# To JPG
pdftoppm -jpeg -r 150 input.pdf output

# Specific pages
pdftoppm -png -f 1 -l 5 input.pdf output

PDF to DOCX:

# Using libreoffice
libreoffice --headless --convert-to docx input.pdf

# Using pandoc
pandoc input.pdf -o output.docx

PDF to Text:

# Simple conversion
pdftotext input.pdf output.txt

# Maintain layout
pdftotext -layout input.pdf output.txt

6. PDF Analysis & Information

Get page count:

pdfinfo document.pdf | grep "Pages:" | awk '{print $2}'

Check PDF version:

pdfinfo document.pdf | grep "PDF version"

Analyze structure:

# Get detailed structure
mutool show input.pdf outline

# Extract fonts
pdffonts input.pdf

7. PDF Optimization

Compress PDF:

# Using ghostscript (screen quality - smallest)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen \
   -dNOPAUSE -dQUIET -dBATCH -sOutputFile=compressed.pdf input.pdf

# Using ghostscript (ebook quality - medium)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/ebook \
   -dNOPAUSE -dQUIET -dBATCH -sOutputFile=compressed.pdf input.pdf

Remove password:

# If you know the password
pdftk secured.pdf input_pw PASSWORD output unsecured.pdf

8. Common Workflows

Extract tables from PDF

# Using tabula-py
tabula-py input.pdf --output-format csv --pages all

# Or use pdfplumber for complex tables

Add watermark

pdftk input.pdf stamp watermark.pdf output watermarked.pdf

Rotate pages

# Rotate all pages 90 degrees clockwise
pdftk input.pdf cat 1-endright output rotated.pdf

# Rotate specific pages
pdftk input.pdf cat 1-5 6right 7-end output rotated.pdf

Tools Required

Make sure these tools are installed:

• poppler-utils (pdftotext, pdfinfo, pdftoppm, pdfunite)
• pdftk or pdftk-java
• ghostscript (gs)
• imagemagick (convert)
• pandoc (for conversions)
• img2pdf (for image to PDF)
• exiftool (for metadata)

Install on Ubuntu/Debian:

sudo apt-get install poppler-utils pdftk ghostscript imagemagick pandoc python3-img2pdf exiftool

Security Notes

• ✅ Always validate PDF file paths before processing
• ✅ Check file sizes to prevent resource exhaustion
• ✅ Sanitize output filenames
• ✅ Be cautious with password-protected PDFs
• ✅ Scan PDFs for malicious content if from untrusted sources

When to Use This Skill

Use /pdf when the user:

• Wants to read or extract text from a PDF
• Needs to create a PDF from other formats
• Wants to merge or split PDFs
• Needs to convert PDFs to images or other formats
• Asks to analyze PDF structure or metadata
• Wants to compress or optimize PDFs

Always confirm destructive operations before executing.