PPT Master

Last reviewed: 2026-04-28

Document Converter 3.4k Stars Open Source

Best for: Converting existing documents — PDF, DOCX, or web URLs — into editable PowerPoint presentations without losing structure

Not ideal for: Creating presentations from scratch when you have no source document; for that, use Anthropic PPTX Skill or PPTAgent v2

Supported Input Sources

📄

PDF

Research papers, reports, whitepapers — layout and structure preserved

📝

DOCX

Word documents, briefs, proposals — headings map to slide titles

🌐

URL

Web articles, blog posts, product pages — content extracted and structured

Output is always native .pptx — all shapes are editable DrawingML, not images or embedded objects.

Core Features

3.4k

GitHub stars

Example outputs

229

Pages of docs

✏️

Native DrawingML Shapes

Every element in the output PPTX is a real, editable PowerPoint shape — not a screenshot or image embed. You can click, resize, restyle, and animate every element in PowerPoint or Keynote.

🗂️

Structure-Aware Conversion

Headings become slide titles, body text maps to content areas, tables convert to real PowerPoint tables, and lists become bullet layouts — not just a dump of text onto blank slides.

📚

15 Example Templates

Ships with 15 real output examples across business, academic, and technical categories. Useful as references for how your converted deck will look.

📖

Comprehensive Documentation

229 pages of documentation covers all conversion scenarios, edge cases, and customization options — the most thoroughly documented skill in this category.

Signals

GitHub Stars

3.4k ⭐

Maintenance

Actively maintained

Output Format

Native .pptx (DrawingML)

License

Open Source

Documentation

229 pages

Example Outputs

15 included

Quality Assessment

Conversion Accuracy

4.2

Editability

4.7

Ease of Use

2.8

Layout Preservation

4.0

Tested: svg_to_pptx.py directly — 7 SVG elements converted to native DrawingML shapes (p:sp), zero image elements (p:pic). Editability claim is accurate. PDF→Markdown pipeline also verified: auto-detects heading hierarchy by font size. Ease of use significantly lower than page implied: this is a 7-step LLM Skill pipeline with 3 blocking user-confirmation nodes (including an "eight-point confirmation" in the Strategist phase) — not a single-command tool. Also requires brew install cairo on macOS for cairosvg support with no fallback guidance.

Install & Use

# Install as a Claude Code Skill
git clone https://github.com/hugohe3/ppt-master
cp -r ppt-master/skills/ppt-master ~/.claude/skills/

# Install Python dependencies
pip install python-pptx svglib reportlab PyMuPDF Pillow requests beautifulsoup4 numpy google-genai openai

# macOS: also needed for SVG rendering
brew install cairo

# Then use naturally in Claude Code:
# "Convert my research paper report.pdf into a presentation"
# "Turn this Word brief into slides I can edit"
# "Make a deck from the content at this URL: ..."
# Note: pipeline includes interactive confirmation steps
        