DavidROliverBA/.claude/skills/pptx-to-page

Skill: PowerPoint to Page

Convert PowerPoint presentations to Obsidian Page notes with text extraction, tables, and images.

When to Use

This skill should be invoked when the user:

• Asks to "convert a PowerPoint to a page note"
• Mentions "/pptx-to-page" explicitly
• Wants to import a presentation into the vault
• Needs to extract slides as images for viewing in Obsidian
• Has a .pptx or .ppt file to process

Parameters

• PPTX Path: Path to the PowerPoint file (can be in Downloads, vault, or anywhere)
• Page Title (optional): Custom title for the Page note (defaults to presentation filename)
• Mode (optional): quick (default) or visual
  • quick: Fast docling extraction with text, tables, embedded images
  • visual: Full slide rendering as PNG images (requires LibreOffice)

Process

Step 1: Locate the PowerPoint File

If the user provides a partial path or just a filename:

1. Check ~/Downloads/ first
2. Check vault +Attachments/ folder
3. Ask user for full path if not found

Step 2: Choose Processing Mode

Ask the user (or infer from context):

Quick Mode (default - recommended for most uses):

• Uses docling for fast text/table extraction
• Uses python-pptx for speaker notes and embedded images
• Processing time: ~1 second for 50 slides
• Best for: searchable content, meeting notes, documentation

Visual Mode (when visual fidelity needed):

• Uses LibreOffice to render full slide images
• Processing time: 1-2 minutes for 50 slides
• Best for: design reviews, exact visual reference

Step 3: Quick Mode Processing (Docling)

from pathlib import Path
from docling.document_converter import DocumentConverter
from pptx import Presentation
import os

def process_pptx_quick(pptx_path, output_dir, title):
    """Quick mode: docling + python-pptx extraction"""

    # 1. Docling for text and tables
    converter = DocumentConverter()
    result = converter.convert(pptx_path)
    doc = result.document

    markdown_content = doc.export_to_markdown()
    tables_count = len(doc.tables) if hasattr(doc, 'tables') else 0
    pictures_count = len(doc.pictures) if hasattr(doc, 'pictures') else 0

    # 2. python-pptx for speaker notes and embedded images
    prs = Presentation(pptx_path)
    speaker_notes = []
    embedded_images = []

    for slide_num, slide in enumerate(prs.slides, 1):
        # Extract speaker notes
        if slide.has_notes_slide:
            notes = slide.notes_slide.notes_text_frame.text.strip()
            if notes:
                speaker_notes.append((slide_num, notes))

        # Extract embedded images
        for shape in slide.shapes:
            if hasattr(shape, "image"):
                img = shape.image
                img_filename = f"{title} - Slide {slide_num:02d} - Image {len(embedded_images)+1}.{img.ext}"
                img_path = output_dir / img_filename
                with open(img_path, "wb") as f:
                    f.write(img.blob)
                embedded_images.append((slide_num, img_filename))

    return {
        'markdown': markdown_content,
        'tables_count': tables_count,
        'pictures_count': pictures_count,
        'speaker_notes': speaker_notes,
        'embedded_images': embedded_images,
        'slide_count': len(prs.slides)
    }

Step 4: Visual Mode Processing (LibreOffice)

Only use if visual mode requested. Check dependencies first:

# Check for LibreOffice (for PPTX to PDF conversion)
which soffice || brew install --cask libreoffice

# Check for poppler (for PDF to image conversion)
which pdftoppm || brew install poppler

# Check for Python libraries
python3 -c "import pdf2image" 2>&1 || pip3 install pdf2image

Then convert PPTX → PDF → PNG images at 200 DPI.

Step 5: Create Page Note

Quick Mode Output Format:

---
type: Page
title: <Presentation Title>
created: YYYY-MM-DD
modified: YYYY-MM-DD
tags:
  - presentation
  - slides
  - pdf-import
sourceType: PPTX
processedWith: docling
slides: <count>
tables: <count>
---

# <Presentation Title>

> Extracted from `<original-filename.pptx>` using docling

**Slides**: <count> | **Tables**: <count> | **Images**: <count>

---

## Content

<docling markdown output - includes headings, text, tables>

---

## Speaker Notes

### Slide 3
<notes content>

### Slide 7
<notes content>

---

## Embedded Images

### Slide 1
![[<title> - Slide 01 - Image 1.png]]

### Slide 4
![[<title> - Slide 04 - Image 1.jpg]]
![[<title> - Slide 04 - Image 2.png]]

---

## Related

- [[<linked meeting note if applicable>]]

Visual Mode Output Format:

---
type: Page
title: <Presentation Title>
created: YYYY-MM-DD
modified: YYYY-MM-DD
tags:
  - presentation
  - slides
sourceType: PPTX
processedWith: libreoffice
slides: <count>
---

# <Presentation Title>

Source: `<original-pptx-filename>`

Total Slides: <count>

---

## Slide 1

![[<title> - Slide 01.png]]

---

## Slide 2

![[<title> - Slide 02.png]]

---

...

Step 6: Link to Context (if applicable)

Ask the user if they want to link this page to:

• A specific meeting note (most common for presentations)
• A project note
• An ADR (if it's architecture-related)

Suggest adding it to the meeting note's "Slides" section.

Mode Selection Guide

| Use Case | Recommended Mode | |----------|-----------------| | Meeting notes reference | Quick | | Technical documentation | Quick | | Searchable archive | Quick | | Design review | Visual | | Exact visual reference needed | Visual | | Large presentations (50+ slides) | Quick | | Offline/slow machine | Quick |

Success Message

Quick Mode:

✅ Page note created: `Page - <title>.md`
📊 Slides processed: <count>
📋 Tables extracted: <count>
🖼️ Embedded images saved: <count> to +Attachments/
💬 Speaker notes captured: <count> slides
⚡ Processing time: <X.XX> seconds
🔗 Link in meeting notes with: `[[Page - <title>]]`

Visual Mode:

✅ Page note created: `Page - <title>.md`
📊 Total slides rendered: <count>
🖼️ Slide images saved to: +Attachments/
⏱️ Processing time: <X> minutes
🔗 Link in meeting notes with: `[[Page - <title>]]`

Error Handling

Common errors and solutions:

1. PPTX not found: Ask user for correct path
2. Docling import error: pip3 install docling
3. python-pptx import error: pip3 install python-pptx
4. LibreOffice missing (visual mode): brew install --cask libreoffice
5. poppler missing (visual mode): brew install poppler
6. Corrupted PPTX: Inform user, suggest opening in PowerPoint first
7. Password-protected: Cannot process, ask user to remove password

Performance Comparison

| Presentation Size | Quick Mode | Visual Mode | |------------------|------------|-------------| | 10 slides | ~0.2 sec | ~20 sec | | 25 slides | ~0.4 sec | ~50 sec | | 50 slides | ~0.8 sec | ~2 min | | 100 slides | ~1.5 sec | ~4 min |

Examples

Example 1: Quick Mode (Default)

User: /pptx-to-page ~/Downloads/Project_Update.pptx

Example 2: Visual Mode

User: /pptx-to-page SAP_Architecture.pptx visual

Example 3: With Meeting Link

User: Convert the Datasphere presentation and add it to today's meeting note

Example 4: With Custom Title

User: Convert SAP_BTP_2026.pptx to a page called "Cloud Platform Architecture 2026"

Integration with Other Skills

This skill works well with:

• /meeting - Add presentation to meeting notes
• /project-status - Include project presentations
• /adr - Attach architecture diagrams from slides
• /weekly-summary - Reference key slides in summaries

Dependencies

Quick Mode (minimal):

• docling - Text and table extraction
• python-pptx - Speaker notes and embedded images

Visual Mode (additional):

• LibreOffice - PPTX to PDF conversion
• poppler (pdftoppm) - PDF to PNG conversion
• pdf2image - Python wrapper for poppler

Notes

• Quick mode is ~100x faster than visual mode
• Quick mode extracts actual text (searchable, copyable)
• Visual mode preserves exact slide appearance
• Always copy the original PPTX to +Attachments/ for reference
• Speaker notes are only available via python-pptx (both modes can include them)
• Consider using quick mode first, then visual mode for specific slides if needed