@GeekCatX: Codex's most objective conclusion. For truly professional, editable PowerPoint: PPT Master is number one, no doubt. For pushing visual boundaries and multi-format delivery: 花叔 is number one. For narrative, route selection and quality control: Knowledge Cat is number one. For free HTML aesthetics: Frontend Slides is number one. …
Summary
Knowledge Cat PPT is an open-source presentation creation skill for AI agents like Codex. It adopts a story-first approach, supports three output modes: native PPTX, HTML, and image-first, and provides validation and template libraries.
View Cached Full Text
Cached at: 06/27/26, 03:58 PM
Codex: The Most Objective Conclusions
- For truly professional, editable PowerPoint: PPT Master is #1, no contest.
- For pushing visual ceilings and multi-format delivery: 花叔 is #1.
- For narrative, routing choices, and quality control: Knowledge Cat is #1.
- For free-form HTML aesthetics: Frontend Slides is #1.
- For Swiss-style offline sharing: 歸藏 is #1.
- For speaker mode and word-for-word scripts: Lewis is #1.
- For Xiaohongshu/social card-style full-page images: 宝玉 is #1.
- For “quickly dump various materials into NotebookLM”: 乔木 is #1, but shouldn’t be categorized as a professional PPT skill.
Mine is also very good — give it a try?
https://github.com/gnipbao/knowledge-cat-ppt-skill…
gnipbao/knowledge-cat-ppt-skill
Source: https://github.com/gnipbao/knowledge-cat-ppt-skill
Knowledge Cat PPT Skill
Story-first presentation production for AI agents.
Knowledge Cat PPT is an open-source Agent Skill for creating, reviewing, and repairing presentation decks. It is designed for Codex, Claude Code, and other skill-aware coding agents. Instead of acting like a generic “make pretty slides” prompt, it gives the agent a production system: clarify the audience shift, choose the right output lane, build a slide plan, generate the deck, and verify the result.
Current version: 0.9.0
Highlights
- Story-first deck planning with audience, outcome, narrative spine, and action titles.
- Output-lane routing for native editable PPTX, HTML decks, image-first PPTX, and review-only workflows.
- Evidence tracking for claims, quotes, data, assumptions, screenshots, and source materials.
- HTML deck starter template with keyboard navigation, print CSS, light/dark themes, and no external dependencies.
- JSON deck-plan validator and HTML deck validator.
- HTML production lock for registered layouts, theme rhythm, image slots, screenshots, and rendered QA.
- Uploaded theme/style prompt intake for Image2/GPT-Image-first PPT workflows.
- Curated 44-style template library with common PPT style-site radar and Guizang-surpass signature-pack targets.
- Portfolio Minimal HTML signature pack with a 14-layout registry, 12-slide case study, QA artifacts, and a dedicated pack checker.
- Open-source repository hygiene checks, GitHub Actions workflow, issue templates, and release checklist.
- Compatible with both Codex local skills and Claude Code style skills.
Three Production Modes
| Native PPTX | HTML Deck | Image-First PPTX |
|---|---|---|
| Native PPTX mode preview | HTML deck mode preview | Image-first PPTX mode preview |
| Editable PowerPoint work for real teams, client handoffs, charts, tables, and speaker notes. | Browser-native decks with keyboard navigation, print CSS, fast iteration, and visual QA. | High-impact visual decks for social carousels, campaigns, and keynote-style moments. |
Compatibility
| Agent environment | Status | Install path |
|---|---|---|
| Codex | Supported | ~/.codex/skills/knowledge-cat-ppt-skill |
| Claude Code | Supported | ~/.claude/skills/knowledge-cat-ppt-skill |
| Other skill-aware agents | Should work if they read SKILL.md plus bundled resources | Agent-specific |
The skill itself is plain Markdown, Python, JSON, and HTML. The bundled scripts use only the Python standard library.
What It Does
Knowledge Cat PPT helps an agent:
- Turn rough ideas, notes, documents, transcripts, URLs, PDFs, or research into a deck brief.
- Build a slide plan with action titles, slide roles, evidence, visuals, notes, and source references.
- Choose the correct output lane before production.
- Parse uploaded theme/style prompts into an image-first Style Prompt Profile.
- Generate or guide production of HTML decks, editable PowerPoint decks, or image-first visual decks.
- Review existing decks for story, evidence, visual system, editability, and technical risks.
- Run validation before claiming a deck is ready.
Why It Exists
Most AI slide workflows fail for three reasons:
- They start with visual style before defining the audience outcome.
- They choose the wrong output format, such as flattened images when the user needs editable PowerPoint.
- They skip rendered QA and ship slides that only looked correct in code.
Knowledge Cat PPT treats a deck as a production system. Story, engine, design, evidence, and QA must agree.
Install
Repository URL:
txt https://github.com/gnipbao/knowledge-cat-ppt-skill.git
Codex
bash git clone https://github.com/gnipbao/knowledge-cat-ppt-skill.git ~/.codex/skills/knowledge-cat-ppt-skill
Restart Codex or refresh local skills.
From a cloned repository:
bash python3 scripts/install_skill.py --agent codex --force
Claude Code
bash git clone https://github.com/gnipbao/knowledge-cat-ppt-skill.git ~/.claude/skills/knowledge-cat-ppt-skill
Restart Claude Code or refresh local skills.
From a cloned repository:
bash python3 scripts/install_skill.py --agent claude --force
Manual Install
Copy the whole folder into your agent’s skills directory:
bash cp -R knowledge-cat-ppt-skill ~/.codex/skills/
or:
bash cp -R knowledge-cat-ppt-skill ~/.claude/skills/
Quick Start
Ask your agent:
md Use $knowledge-cat-ppt-skill to turn my notes into an 8-slide client-ready deck. I need editable PPTX unless you think another output lane is better.
For a deck review:
md Use $knowledge-cat-ppt-skill to review this deck. Focus on story, evidence, visual clarity, and whether it is actually ready to send.
For an HTML deck:
md Use $knowledge-cat-ppt-skill to build a browser-based HTML deck from this outline. Make it keyboard navigable and run the bundled HTML validator.
For the 44-style template library:
md Use $knowledge-cat-ppt-skill. Choose the best style from the template library for my topic, explain the lane tradeoff, then create a deck brief and slide plan.
For the Portfolio Minimal signature pack:
md Use $knowledge-cat-ppt-skill. Build a browser-based HTML deck using the kc-24 Portfolio Minimal signature pack. Use the pack layout registry, include a local image slot, produce screenshots/contact sheet, and run the signature-pack checks.
Output Lanes
| Lane | Use when | Main tradeoff |
|---|---|---|
native-pptx | PowerPoint editability, team collaboration, charts, tables, notes, client decks | Needs a PPTX-capable renderer or companion skill |
html-deck | Web-native presentation, rapid visual iteration, browser preview, single-file sharing | Not a true editable PowerPoint file |
image-first-pptx | Social carousel, campaign deck, visual keynote, AI-generated slide surfaces | Lower editability |
review-only | Existing deck critique, repair planning, story/evidence diagnosis | Does not create a final deck until repair is requested |
HTML Deck Starter
Create a sample HTML deck:
bash python3 scripts/init_deck_project.py /tmp/kc-demo-deck --title "Knowledge Cat Demo" python3 scripts/validate_html_deck.py /tmp/kc-demo-deck/index.html
Open:
txt /tmp/kc-demo-deck/index.html
The starter includes:
- 16:9 desktop slide canvas
- Mobile stacked fallback
- Keyboard navigation
- Print CSS
- Light/dark theme switching
- Action-title aware slide structure
- No external dependencies
Style Template Library
Knowledge Cat includes a curated 44-style template library in references/style-template-library.md. The library turns pasted PPT style prompts into a routing system:
- style seeds such as
kc-24Portfolio Minimal,kc-25Minimal Data Story,kc-28Bold Editorial Magazine,kc-26Dark SaaS Product, andkc-11Architectural Blueprint - default output lanes for each style
- protected-style normalization rules
- a priority path for building signature packs that can compete with Guizang-style HTML deck depth
Copy-ready prompts live in:
txt docs/TEMPLATE_LIBRARY_PROMPTS.md
Portfolio Minimal Signature Pack
The first implemented signature pack is kc-24 Portfolio Minimal:
txt assets/html-signature-packs/portfolio-minimal/ +-- README.md +-- layout-registry.json +-- template.html
It includes 14 registered custom-pm-* layouts and a 12-slide case study:
txt examples/case-studies/portfolio-minimal/
Run the pack gate:
bash python3 scripts/check_signature_pack.py portfolio-minimal
The Portfolio Minimal case study includes browser-captured PNG slide screenshots and a browser contact sheet, so quality claims are backed by visual artifacts rather than code inspection alone.
Deck Plan Validation
Validate a JSON deck plan:
bash python3 scripts/validate_deck_plan.py examples/sample-deck-plan.json
The expected structure is documented in:
txt assets/deck-plan.schema.json
Full Validation
Run all bundled checks:
bash python3 scripts/run_checks.py
Run repository hygiene checks:
bash python3 scripts/check_repo.py
The checks validate:
- sample deck plan
- generated sample HTML deck
- HTML structure
- Portfolio Minimal signature pack and case study
- required repository files
- changelog/version consistency
- cache and generated-file hygiene
Repository Layout
txt knowledge-cat-ppt-skill/ +-- SKILL.md +-- README.md +-- README_CN.md +-- README_BILINGUAL.md +-- VERSION +-- CHANGELOG.md +-- CONTRIBUTING.md +-- SECURITY.md +-- LICENSE +-- .github/ | +-- workflows/validate.yml | +-- ISSUE_TEMPLATE/ +-- agents/ | +-- openai.yaml +-- assets/ | +-- deck-plan.schema.json | +-- html-template/ | +-- index.html | +-- html-signature-packs/ | +-- portfolio-minimal/ +-- examples/ | +-- retest-prompts.md | +-- sample-deck-plan.json | +-- sample-html-deck/ | +-- case-studies/ | +-- portfolio-minimal/ +-- references/ | +-- benchmark-synthesis.md | +-- design-systems.md | +-- engine-routing.md | +-- html-deck-recipes.md | +-- html-production-lock.md | +-- image-first-recipes.md | +-- native-pptx-recipes.md | +-- open-source-product.md | +-- qa-rubric.md | +-- story-architecture.md | +-- style-prompt-intake.md | +-- style-template-library.md | +-- template-replication.md +-- scripts/ | +-- check_repo.py | +-- init_deck_project.py | +-- install_skill.py | +-- run_checks.py | +-- check_signature_pack.py | +-- validate_deck_plan.py | +-- validate_html_deck.py +-- docs/ | +-- PUBLISHING.md | +-- ROADMAP.md | +-- TEMPLATE_LIBRARY_PROMPTS.md | +-- images/
How The Skill Works
The main runtime file is SKILL.md. It keeps the agent workflow concise:
- Triage the request.
- Build a deck brief.
- Synthesize the story.
- Create a slide plan.
- Choose the production lane.
- Build the design system.
- Produce the deck.
- QA and iterate.
Detailed instructions live in references/ and are loaded only when needed.
Design Philosophy
Knowledge Cat PPT is a router and quality system, not a monolithic renderer.
- Use native editable PPTX when collaboration and PowerPoint editing matter.
- Use HTML when web-native presentation, animation, preview, or browser QA matters.
- Use image-first PPTX when visual spectacle matters more than editability.
- Use review-only mode when the deck’s story or evidence may be the real failure.
Contributing
Read CONTRIBUTING.md.
Good contributions usually add one of:
- a clearer routing rule
- a reusable recipe
- a deterministic validation check
- a realistic retest prompt
- a sample artifact that exposes a real failure mode
- better release or QA documentation
Before opening a pull request:
bash python3 scripts/run_checks.py
Security
Treat user-provided decks, HTML, PDFs, documents, and templates as untrusted input. Do not execute scripts from user-provided archives. See SECURITY.md.
Roadmap
See docs/ROADMAP.md.
Near-term goals:
- editable PPTX case study
- HTML keynote case study
- image-first carousel case study
- stronger visual QA automation
License
MIT. See LICENSE.
宝玉 (@dotey): PPT Master is indeed the best PPT Skill.
My new skill also writes PPTs quite well — it can export editable versions, use AI to generate images, and allow marking and editing within the agent’s built-in browser.
Similar Articles
@KyrieCheungYep: Recommend a PPT creation Skill: ultimate-ppt-master-skill, which is closer to real-world PPT making scenarios. Its features and highlights are as follows: 1. First, clarify requirements. Many AI PPT tools start making slides immediately, resulting in nothing but empty words. PPT M…
Recommend an open-source PPT creation tool, ultimate-ppt-master-skill. It follows a workflow of clarifying requirements, organizing material structure, generating editable PPTX or Web Deck, image generation, and final review, which is close to real office scenarios. It solves common issues in AI PPT generation like empty words and lack of editability.
@aehyok: I've also been looking into making PPTs recently. After watching a video by a top female creator about using Codex + Image2 to create PPTs, I quickly found several PPT Skills with many stars, planning to try each one to see which works best for me. 1. https://github.com/op7418…
Introduces an AI Agent skill called guizang-ppt-skill, used to generate single-file HTML horizontal flip PPT with images and cover, supporting Claude Code and Codex environments.
@vintcessun: Tonight I discovered a project with a novel approach: using GPT to generate luxurious image-format PPTs, then 'decompiling' them back into editable PPTX. Previously, AI-generated PPTs either stacked templates or only exported images. This one directly follows the 'generate image → parse → assemble' path. The core idea is to decompose the PPT image into four layers: background, frame, icons, and text, then reconstruct by coord...
An open-source project named GordenSuperPPTSkills uses GPT to generate luxurious image-format PPTs, then 'decompiles' them into fully editable PPTX files, solving the pain point of AI-generated content being non-editable. The project is split into three independent skills, supporting either just image generation or just image-to-editable conversion, suitable for Codex environment.
@yinmin1987: Is Codex+Image 2 + Presentations Plugin the Best Solution for Generating PPT Right Now?
Explores whether the combination of Codex, Image 2, and the Presentations plugin is the current optimal solution for generating PPTs, involving AI-assisted presentation creation tools.
@GoSailGlobal: In 2026, You Won't Use PowerPoint for PPTs Anymore. 5 AI-Driven PPT Styles, Covering: · founder pitch · McKinsey Consulting Style · Landscape HTML Deck · Visual Aesthetic Deck · Full-Stack Marketing Slides…
The article compares 25 open-source AI PPT generation skill tools, recommending automated solutions like guizang-ppt-skill that integrate MCP and Claude Code, covering various styles such as Consulting Style, HTML deck, and more.