PPT Master is an open-source Python tool that turns PDFs, DOCX files, URLs or Markdown into fully editable PowerPoint decks. You install Python 3.10+, grab the repo from GitHub or AtomGit, run pip install -r requirements.txt, then feed your source material into an AI IDE—Claude Code, VS Code + Copilot, Cursor or Codebuddy. In 10–20 minutes of chat prompts, the AI uses a predefined workflow to analyze your content, generate DrawingML shapes, text boxes and charts, then export a native .pptx. The author reports each deck costs about $0.08 in Copilot tokens and runs entirely on your machine except for API calls to the AI model. The project emphasizes four core advantages over competing services. First, every element in the output is a real, clickable PowerPoint object; no image imports or locked web snapshots. Second, it’s free and transparent: you pay only for your AI subscription and control your spending. Third, data stays local—no forced uploads to third-party servers. Fourth, you avoid vendor lock-in: support for multiple IDEs and models including Claude, GPT, Gemini, Kimi and others. Setup takes about 15 minutes on a fresh machine. Windows users follow an extra PATH and execution-policy guide, macOS and Linux users install Python via Homebrew or apt and go straight to pip. Two optional fallbacks—Node.js for certain WeChat imports and Pandoc for legacy formats—cover edge cases. Once installed, you drop your source into the projects/ folder or paste text into the IDE chat, confirm design specs (template, page count, aspect ratio) and let the AI handle content analysis, layout, SVG generation and export. Exports land in exports/ as both a fully editable shapes-based PPTX and a snapshot version for reference. The author is a finance professional (CPA, CPV, Consulting Engineer) who built PPT Master out of frustration with image-only slides. He maintains the code solo, funds AI-model bills himself, and updates docs, templates and bug fixes through the same workflow. Sponsorships cover ongoing work; paid consulting is available for custom templates or enterprise deployments.