PPT Master

repository·main·Indexed 12 days ago

https://github.com/hugohe3/ppt-master

An AI-powered engine that transforms documents (PDF, DOCX, web pages) into natively editable PowerPoint (.pptx) presentations. Unlike tools that produce static images, PPT Master generates real PowerPoint objects, including slide masters, data-backed charts, tables, and native animations. Designed to run locally within agent-capable AI tools like Claude, GPT, and Gemini, it supports custom brand templates, voice cloning for narration, and flexible image acquisition via AI generation or web search.

Tokens
338.3K
Snippets
541
Records
1.2K
Agent score
100%

What's inside PPT Master

  1. Overview of PPT Master

    main

    PPT Master is an AI-driven tool that generates native, fully editable PowerPoint (.pptx) presentations from various source materials such as PDFs, DOCX files, or web pages. Unlike tools that produce static layouts or flat text boxes, PPT Master creates decks using real PowerPoint primitives, including:

    • Native Slide Masters: Ensures consistent styling and structure.
    • Native Shapes and Objects: Uses actual PowerPoint shapes, not just images.
    • Data-backed Charts and Tables: Generates real PowerPoint chart and table objects that can be edited.
    • Advanced Animations: Supports native slide transitions and opt-in animations (entrance, emphasis, motion-path, and exit).
    • Speaker Notes: Generates notes that can be converted into audio narration or video.

    The workflow is designed to run inside agent-capable AI tools, allowing you to provide a topic or material and have the AI generate the deck locally on your machine, ensuring data privacy and avoiding platform lock-in.

  2. Overview of the Science Explainer style

    main

    The science-explainer style is a design specification intended for public-facing science and technology communication. Its primary goal is to move an audience from curiosity to genuine understanding of a single idea using visual analogies and instructional argument flows, without sacrificing scientific accuracy for accessibility.

    Best Use Cases:

    • Public lectures and museum outreach.
    • Popular-science and technology explainers.
    • Journalist and policy briefings.
    • Internal cross-discipline explanations.

    Core Philosophy:

    • Instructional Mode: Build understanding step-by-step from familiar ground.
    • One Idea Per Page: Each page should carry exactly one idea and be understandable in seconds.
    • Analogy Discipline: Always mark analogies as such and explicitly state where the analogy stops being true.
    style_id: science-explainer
    kind: style
    summary: Public-explanation method that builds understanding from familiar ground through visual analogy, without trading accuracy for accessibility.
    keywords: [explainer, science-communication, popular, analogy, public]
  3. Understand the Operating Review style specification

    main

    The operating-review style is a specialized design and communication method for recurring business reviews (monthly/quarterly). It is designed to separate results, variance, causes, and commitments without softening bad news. It focuses on high-density, data-driven reporting where the audience can compare different time periods easily because the page structure remains stable.

    Key Characteristics:

    • Best Fit: Business reviews, operational performance readouts, and metric reporting.
    • Communication Mode: A hybrid of 'briefing' (stating results plainly) and 'pyramid' (resolving variances by showing deviation $\rightarrow$ cause $\rightarrow$ evidence $\rightarrow$ response).
    • Visual Style: data-journalism with a focus on native, editable charts and tables rather than flattened images.
    • Core Discipline: Never title a page with a metric name alone if the page is meant to explain a movement; always state the result and its direction against plan in the title.
    style_id: operating-review
  4. Understand the core capabilities of PPT Master

    main

    PPT Master is an open-source tool designed to generate high-fidelity, native PowerPoint (.pptx) documents from various source documents using AI. Unlike many AI presentation tools that produce flattened images or simple editable skins, PPT Master authors the PowerPoint native object model directly.

    Key capabilities include:

    • Native PowerPoint Objects: Generates real slide masters (p:sldMaster), layouts (p:sldLayout), native shapes (with adjustment handles), connectors, and freeform paths.
    • Data-Backed Elements: Supports native charts and tables (optional) and full text/picture/fill/effect models (gradients, shadows, glow, etc.).
    • Animations & Narration: Creates real OOXML transitions, entrance/exit animations, and converts speaker notes into voice narration.
    • Template Distillation: Can extract and reuse brand styles and layouts from reference materials.
    • Narrative Intelligence: Uses a 'Strategist' layer to determine the logical structure (e.g., pyramid, narrative, instructional) before designing visuals.
    • Multi-Format Output: Supports various aspect ratios including 16:9, 4:3, 3:4 (Xiaohongshu), 1:1 (WeChat/Instagram), 9:16 (Stories), and A4 print.
  5. What PPT Master is not (Non-goals)

    main

    To understand the scope of PPT Master, note that it is not intended to be:

    • A zero-setup, instant-slide browser SaaS.
    • A fully autonomous replacement for human presentation judgment.
    • A general office assistant, research platform, or image/audio provider.
    • A complete PowerPoint clone or real-time collaboration service.
    • An arbitrary SVG-to-PPTX or OOXML conversion service.
    • A system that reconstructs missing Master/Layout intent from finished files.
    • An in-place upgrader that grafts inferred templates onto existing files.
    • A speed-first product that sacrifices deck coherence or native editability.
  6. Identify what PPT Master does NOT do (Non-goals)

    main

    To avoid proposing features that are intentionally out of scope, note the following non-goals:

    • No arbitrary PPTX placeholder refilling: The tool does not perform unreviewed, schema-free substitution against third-party placeholder systems. It only supports template-fill-pptx (refilling selected slides) or create-template (deriving workspaces from source evidence).
    • No native charts by default: Charts are SVG by design to maintain cross-renderer consistency. Native charts are only available via explicit opt-in.
    • No standalone form factors: The project is designed as a chat-driven workflow/skill for AI agents (e.g., Claude Code, Cursor, VS Code agents). It will not become a standalone CLI, SaaS, or desktop app.
    • No pure speed optimization at the cost of quality: The project prioritizes high-quality output (approx. 20 minutes for a high-quality PPTX) over raw speed.
  7. Navigate PPT Master documentation

    main

    The project documentation is organized into several key areas:

    • Getting Started: Three-step process for creating decks, including templates, previews, and animations.
    • Decision Support: Guides on why to choose PPT Master and understanding its capability boundaries.
    • Installation: Specific instructions for Windows users.
    • Core Logic: SKILL.md contains core processes and rules.
    • Formats: canvas-formats.md details supported formats like PPT 16:9, Xiaohongshu, and WeChat Moments.
    • Tools: scripts/README.md lists all available scripts and commands.
    • Examples: examples/README.md contains project examples.
    • Technical Design: technical-design.md covers architecture and the use of SVG.
    • Troubleshooting: faq.md covers model selection, costs, layout issues, and custom templates.
  8. Core commitments and constraints of PPT Master

    main

    PPT Master operates under three primary commitments:

    1. Cost Transparency: The tool is open-source and free. The only cost incurred is your own AI model usage fees.
    2. Data Privacy: The entire process is completed on your local computer, except for the dialogue with the AI model.
    3. Platform Agnostic: It can be driven by any AI IDE or tool with agent capabilities (e.g., Claude, GPT, Gemini, Kimi).
  9. Understand the PPTX Animation Core ownership and responsibilities

    main

    The PPTX Animation Core manages the object-effect vocabulary, trigger semantics, OOXML timing writer, semantic read-back, and package validation for PowerPoint OOXML.

    Key Ownership Mapping:

    • Effect registry, timing writer, and read-back: scripts/pptx_animations.py
    • Sidecar parsing and SVG target discovery: svg_to_pptx/animation_config.py
    • SVG group-to-shape mapping: svg_to_pptx/drawingml/converter.py
    • Generated PPTX resolution and validation: svg_to_pptx/pptx_package/builder.py
    • Narration timing merge: svg_to_pptx/pptx_package/narration.py
    • Public authoring contract: references/animations.md
    • Customization stage: workflows/stages/customize-animations.md

    Important Constraint: Only the generated SVG-to-PPTX route writes object animations. Direct-PPTX routes preserve existing source animations and perform structural package validation but do not author or resolve new animation effects.

  10. Understand the PPT Master workflow and capabilities

    main

    PPT Master is a workflow designed to run inside agent-capable AI tools (like Claude, GPT, Gemini, or Kimi) to generate natively editable .pptx files. Unlike tools that export flat images, PPT Master delivers PowerPoint's native object model, including:

    • Native Shapes and Connectors: Elements with working adjustment handles.
    • Data-backed Content: Charts and tables generated on demand.
    • Full Styling: Support for the complete text, picture, fill, and effect model.
    • Slide Masters: Support for p:sldMaster and p:sldLayout inheritance via template/structured routes.

    Available Workflows

    • New Deck Generation: Creating a deck from source documents (e.g., "make a deck from this PDF").
    • Template Distillation: Extracting reusable brand, style, layout, or deck templates from references.
    • Content Filling: Adding new content to an existing .pptx while preserving its original design.
    • Enhancement: Adding native transitions, animations, and narration to finished decks.
  11. Explore PPT Master Toolset Documentation

    main

    For more detailed information on specific components of the toolset, refer to the following documentation guides:

    • Conversion Tools: ./docs/conversion.md
    • Project Tools: ./docs/project.md
    • SVG Pipeline Tools: ./docs/svg-pipeline.md
    • PPTX Transition Core: ./docs/pptx-transitions.md
    • Image Tools: ./docs/image.md
    • Troubleshooting: ./docs/troubleshooting.md
    • Skill Entry: ../SKILL.md
  12. Design Specification for 中汽研 (CATARC) Deck

    main

    The 中汽研 (CATARC) deck is a professional engineering-style template designed for product certification displays, evaluation reports, and technical exchanges. It uses a deep blue and white visual system to establish credibility and professional order.

    Core Visual Identity

    • Visual Style: Deep blue backgrounds, white content pages, square chapter numbering, and clean grids.
    • Master Families: The template is split into two distinct Master families rather than individual layouts:
      • CATARC Dark: Used for covers, chapter headers, and closing slides (deep blue background).
      • CATARC Light: Used for agendas and body content (white background).

    Technical Canvas Settings

    • Format: ppt169 (16:9 aspect ratio)
    • Dimensions: 1280 x 720
    • Viewbox: 0 0 1280 720
    ---
    deck_id: 中汽研
    kind: deck
    category: brand
    summary: 中汽研认证展示、评测汇报与技术交流,用于建立可信理解并推进评审或合作;采用专业的深蓝工程视觉。
    keywords: [中汽研, 认证, 评测, 技术交流]
    primary_color: "#004098"
    canvas_format: ppt169
    canvas_width: 1280
    canvas_height: 720
    canvas_viewbox: "0 0 1280 720"
    source_canvas_width: 1280
    source_canvas_height: 720
    source_viewbox: "0 0 1280 720"
    replication_mode: fidelity
    native_structure_mode: structured
    page_count: 5
    ---