extract-pptx-content-and-structure

name	Extract PPTX Content and Structure
description	Use this skill to extract text, slide titles, and content from PowerPoint presentations to accurately determine their subject matter for classification.

When to use this skill

• When you need to read the actual content from .pptx files
• Before sorting presentations into subject folders
• When filenames don't clearly indicate the presentation topic

How to extract PPTX content

Using markitdown for structured extraction

python -m markitdown input.pptx > output.md

Using Python with python-pptx

from pptx import Presentation

def extract_pptx_content(pptx_path):
    prs = Presentation(pptx_path)
    content = []
    
    for slide_num, slide in enumerate(prs.slides, 1):
        slide_text = f"--- SLIDE {slide_num} ---\n"
        
        for shape in slide.shapes:
            if hasattr(shape, "text"):
                slide_text += shape.text + "\n"
        
        content.append(slide_text)
    
    return "\n".join(content)

Key information to extract

1. Title slide: Main topic of presentation
2. Slide titles: Each slide's heading indicates content areas
3. Body text: Bullet points and paragraphs contain subject keywords
4. First 5 slides: Usually establish the presentation's main subject

Subject indicators to look for

• Look for the same keyword patterns as PDF extraction
• Note technical terms, proper nouns, and repeated concepts
• Section headers often reveal the subject area

Store extracted content

Keep the full extracted text for analysis during subject classification.