Magazine Web Ppt logo

Magazine Web Ppt

OrganizationPopular
nexu-io
magazine-web-ppt

For marketing and gtm work: bind launches, campaigns, events, and brand plans to growth and pipeline outcomes. Built around the core query "annual-marketing-plan", with GTM strategy lead judgment, buyer-ready proof, and this outcome: approve launch plan, campaign budget, or GTM motion.

Overview

Publishernexu-io
Repositoryopen-design
Skill namemagazine-web-ppt
Stars
96.7K
Forks
11.2K
Bundled files
9
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 9 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by nexu-io on GitHub. Read the source before you install it.

Installation

Install the Magazine Web Ppt AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/nexu-io/open-design.git /tmp/open-design
mkdir -p .claude/skills
cp -r /tmp/open-design/design-templates/guizang-ppt .claude/skills/magazine-web-ppt
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Magazine Web Ppt in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Magazine Web Ppt on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Magazine Web Ppt is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Magazine Web Ppt

What this skill does

Generates a single-file HTML horizontal-swipe deck with a visual tone of:

  • A electronic magazine + electronic ink hybrid style
  • A WebGL fluid / contour / dispersion background (visible on hero pages)
  • Serif headlines (Noto Serif SC + Playfair Display) + sans-serif body (Noto Sans SC + Inter) + monospace metadata (IBM Plex Mono)
  • Lucide line icons (no emoji)
  • Horizontal left/right paging (keyboard ← →, scroll wheel, touch swipe, bottom dots, ESC index)
  • Smooth theme interpolation: when you flip to a hero page, the colors and shader transition softly

This skill's aesthetic is not "corporate PPT," nor "consumer-internet UI": it looks like Monocle magazine after code was stitched onto it.

When to use

Suitable scenarios:

  • Offline talks / internal industry talks / private salons
  • AI new-product launches / demo day
  • Talks with a strong personal style
  • A web version of slides that you "make once, with no paging tool needed"

Unsuitable scenarios:

  • Long stretches of tabular data, stacked charts (use a regular PPT)
  • Training courseware (information density too low)
  • Multi-person collaborative editing (this is static HTML)

Workflow

Step 0 · Infer a direction; ask only when comparison is requested

Infer one of the 5 magazine directions from the brief, audience, content, and known context. Each direction packages its theme color, recommended layouts, chrome style, and slide count. Use the closest fit and continue without asking.

Only when the user explicitly asks to see or compare direction options, open references/styles.md and present the 1-line summaries in one direction-cards question form:

1. Monocle Editorial · International magazine style ✦ default
2. WIRED Tech · Data + engineering
3. Kinfolk Slow · Slow living / humanist
4. Domus Architectural · Architecture / sense of space
5. Lab / Reference · Academic + craft manual

If the user says "I don't know, you recommend" or provides no direction cue: default to Monocle Editorial without asking, because it has the lowest failure probability. If the user mentions "AI / benchmark / technical launch": choose WIRED; "reading / private / social circle": choose Kinfolk; "design / architecture / portfolio": choose Domus; "research / academic / methodology": choose Lab.

After picking a direction, create or update 项目记录.md (Project Record) in the project folder, with the first line clearly stating direction + theme color + audience + duration (see the template at the end of styles.md). Do not change direction at any point: switching midway = everything before is wasted.

Step 1 · Resolve intent

If the user has already given a complete outline + images, you can skip straight to Step 2.

Infer the checklist below from the brief, project metadata, supplied content, and conversation. If a missing answer would materially change the deck and no safe default exists, ask only those unresolved fields in one consolidated question form. Do not ask the checklist one by one.

6-question clarifying checklist

Direction and theme color are inferred in Step 0 unless the user explicitly asks to compare options.

#QuestionWhy ask it
1Who is the audience? Sharing scenario? (internal industry / commercial launch / demo day / private salon)determines the language style and depth
2Talk duration?15 min ≈ 10 pages, 30 min ≈ 20 pages, 45 min ≈ 25-30 pages (see each direction's recommended range in styles.md)
3Any source material? (documents / data / old PPT / article links)with material, build on it; without, help them build
4Any images? Where are they?see the "Image convention" below
5Which theme color do you want?✓ already decided by the direction in Step 0
6Any hard constraints? (must include XX data / must not show YY)avoids rework
Outline assistance (if the user has no outline)

Use the "narrative arc" template to build the skeleton, then fill in content:

Hook       → 1 page   : throw out a contrast / question / hard number that makes people stop
Context    → 1-2 pages : explain the background / who you are / why this topic
Core       → 3-5 pages : the core content, interleaving Layouts 4/5/6/9/10
Shift      → 1 page    : break the expectation / propose a new view
Takeaway   → 1-2 pages : a punchline / a suspenseful question / a call to action

Narrative arc + page-count plan + theme rhythm table (see layouts.md); align all three tables before moving to Step 2.

It's best to save the outline as 项目记录.md (Project Record) or 大纲-v1.md (Outline v1) for later iteration.

Image convention (tell the user)

Make these clear to the user before starting:

  • Folder location: under 项目/XXX/ppt/images/ (sibling to index.html)
  • Naming convention: {page number}-{semantic}.{ext}, e.g. 01-cover.jpg / 03-figma.jpg / 05-dashboard.png
    • Pad the page number with zeros for easy sorting
    • Use English for the semantic part, short, specific, matching the content
  • Spec recommendations:
    • Each image ≥ 1600px wide (avoid blur on large screens)
    • JPG for photos/screenshots, PNG for transparent UI/charts
    • Keep total size under 10MB (affects paging smoothness)
  • How to replace: same-name overwrite is the most stable (no need to change the path in the HTML); if the filename changes, remember to globally search images/oldname and change it to the new name
  • What if there are no images: align with the user; you can generate the structure with placeholder color blocks first and fill in the images later, but tell them that image-text mix pages like layouts 4/5/10 can't have their visuals verified without images

Step 2 · Copy the template

Copy assets/template.html to the target location (usually 项目/XXX/ppt/index.html), and create an images/ folder as a sibling, ready to receive images.

bash
mkdir -p "项目/XXX/ppt/images"
cp "<SKILL_ROOT>/assets/template.html" "项目/XXX/ppt/index.html"

template.html is a complete, runnable file: CSS, WebGL shader, paging JS, font/icon CDNs are all preset; only inside <main id="deck"> are there 3 example slides (cover, act divider, blank filler page).

2.1 · Placeholders you must change (easy to miss)

Right after copying, change the placeholders below, or the browser tab will show awkward text like "[必填] replace with the PPT title":

LocationOriginalChange to
<title>[必填] 替换为 PPT 标题 · Deck Titlethe actual deck title (e.g. 一种新的工作方式 · Luke Wroblewski)

The first thing to do every time you finish copying template.html: grep for "[必填]" to confirm everything is replaced.

2.2 · Choose the theme color (5 presets · no customization allowed)

This skill only allows picking one of 5 carefully tuned presets; it does not accept user-defined hex values: a wrong color pairing makes the visuals ugly in an instant, and protecting the aesthetic matters more than offering freedom.

#ThemeBest for
1🖋 Ink Classicgeneral / commercial launch / the default when you don't know what to pick
2🌊 Indigo Porcelaintech / research / data / technical keynotes
3🌿 Forest Inknature / sustainability / culture / non-fiction
4🍂 Kraft Papernostalgic / humanist / literary / indie magazines
5🌙 Duneart / design / creative / gallery

How to do it:

  1. Recommend one based on the content theme, or just ask the user which one to pick
  2. Open references/themes.md and find the :root block for the corresponding theme
  3. Replace as a whole the lines marked with the "theme color" comment in the opening :root{ block of assets/template.html (the copied version): --ink / --ink-rgb / --paper / --paper-rgb / --paper-tint / --ink-tint
  4. All other CSS flows through var(--...), no other changes needed

Hard rules:

  • A deck uses only one theme, don't change colors midway
  • Don't accept an arbitrary hex value from the user: gently decline and show the 5 presets to pick from
  • Don't mix and match (e.g. take ink from Ink Classic and paper from Dune): it clashes completely

Step 3 · Fill in content

3.0 · Pre-check: class names must be defined in template.html (most important)

This is the source of all generation problems. The layouts.md skeletons use many class names (h-hero / h-xl / stat-card / pipeline / grid-2-7-5, etc.); if assets/template.html's <style> has no corresponding definition, the browser falls back to default styles: the big title turns sans-serif, the data cards squeeze into a clump, the pipeline mushes into one line, the images pile up at the bottom of the page.

Before writing any slide code:

  1. First Read assets/template.html (at least through the end of the <style> block)
  2. Against layouts.md's Pre-flight list, confirm that every class you're going to use exists in the <style>
  3. If a class is missing: add it to template.html's <style>, don't rewrite it inline on every slide
  4. template.html is the single source of class names: don't invent new class names; if you need a custom one, use style="..." inline

Classes commonly missed (must be confirmed to exist in advance): h-hero / h-xl / h-sub / h-md / lead / kicker / meta-row / stat-card / stat-label / stat-nb / stat-unit / stat-note / pipeline-section / pipeline-label / pipeline / step / step-nb / step-title / step-desc / grid-2-7-5 / grid-2-6-6 / grid-2-8-4 / grid-3-3 / grid-6 / grid-3 / grid-4 / frame / frame-img / img-cap / callout / callout-src / chrome / foot

3.0.5 · Plan the theme rhythm (equally important as the class pre-check)

Before picking layouts, you must first list the theme class for each page (hero dark / hero light / light / dark) and write it to a document or draft to align. For detailed rules, see the "Theme rhythm planning" section at the top of references/layouts.md.

Mandatory rules:

  • Every page section must carry one of light / dark / hero light / hero dark; don't write just hero
  • 3+ consecutive pages on the same theme = visual fatigue, not allowed
  • 8+ pages must have ≥1 hero dark + ≥1 hero light
  • The whole deck cannot have only light body pages; it must have a dark body page to create breathing room
  • Insert 1 hero page (cover/divider/question/big quote) every 3-4 pages

Self-check after generating: grep 'class="slide' index.html lists all themes; manually confirm the rhythm is reasonable before delivering.

3.1 · Pick a layout

Don't write slides from scratch. Open references/layouts.md; it has 10 ready-made layout skeletons, each a complete, paste-ready <section> code block:

LayoutUse
1. Coverpage 1
2. Act dividereach act's opening
3. Big numbersthrow out hard data
4. Lead image + text (Quote + Image)identity contrast / story
5. Image gridmulti-image comparison / screenshot evidence
6. Two-column pipeline (Pipeline)workflow process
7. Suspenseful close / question pageact close / wrap-up
8. Big quote page (Big Quote)serif punchline / takeaway
9. Side-by-side comparison (Before / After)old model vs new model
10. Image + text mix (Lead Image + Side Text)information-dense image-text pages

Pick the corresponding layout, paste it in, change the copy and image paths. Be sure to complete the 3.0 pre-check first.

3.2 · Image ratio rules

Always use a standard ratio, never the source image's odd ratio (like 2592/1798):

ScenarioRecommended ratio
Lead image + text, main image16:10 or 4:3 + max-height:56vh
Image grid (multi-image comparison)fixed height:26vh, not aspect-ratio
Small image left + text right1:1 or 3:2
Full-screen hero visual16:9 + max-height:64vh
Image + text mix, small inset3:2 or 3:4

Never use align-self:end on an image: it slides to the bottom of the cell, hidden by the browser toolbar. Use a grid container + align-items:start (already preset in the template) so the image sticks to the top; if you want the left column to stick to the bottom, use flex column + justify-content:space-between.

Component details (fonts, colors, grids, icons, callout, stat-card, etc.) are in references/components.md.

Step 4 · Self-check against the checklist

After generating, always open references/checklist.md and check item by item. It summarizes every pitfall stepped on during real iteration; the P0-level issues (emoji, images breaking out, title wrapping, type division of labor) must all pass.

A few items to watch especially:

  1. Big titles must be serif: if it shows as sans-serif, 99% of the time the Step 3.0 pre-check wasn't done and the h-hero class is missing from template.html
  2. In image grids use only height:Nvh, not aspect-ratio (it breaks out)
  3. Images must not pile up at the bottom of the page: don't use align-self:end, use grid + align-items:start (see Step 3.2)
  4. Images may only use standard ratios (16:10 / 4:3 / 3:2 / 1:1 / 16:9), don't copy the source image's odd ratio
  5. Chinese big titles ≤ 5 characters and nowrap (avoid 1 character per line)
  6. Use Lucide, not emoji
  7. Serif for titles, sans-serif for body, mono for metadata

Step 5 · Local preview

Just open index.html directly in the browser. On macOS:

bash
open "项目/XXX/ppt/index.html"

No local server needed. Images use the relative path images/xxx.png.

Step 6 · Iterate

Modify based on user feedback: the template's CSS is highly parameterized, and 90% of adjustments are changing inline styles (font size font-size:Xvw / height height:Yvh / spacing gap:Zvh).


Resource file tour

magazine-web-ppt/
├── SKILL.md              ← you are reading this
├── assets/
│   ├── template.html     ← the complete runnable template (seed file)
│   └── example-slides.html ← a 9-page sample deck (for the Examples preview)
└── references/
    ├── styles.md         ← the 5 magazine directions (Monocle / WIRED / Kinfolk / Domus / Lab)
    ├── components.md     ← component handbook (fonts, color, grid, icons, callout, stat, pipeline...)
    ├── layouts.md        ← 10 page layout skeletons (paste-ready)
    ├── themes.md         ← 5 theme color presets (pick only, no customization)
    └── checklist.md      ← quality checklist (P0/P1/P2/P3 tiers)

Recommended loading order:

  1. First read through SKILL.md (this file) to understand the whole
  2. When picking a direction in Step 0, read styles.md: each of the 5 directions packages up theme color + recommended layouts + chrome style
  3. After clarifying intent in Step 1, if the direction needs confirmation, read themes.md for palette details
  4. Before starting, Read the <style> block of assets/template.html: this is the single source of class names; missing classes crash the whole page's styling
  5. Read layouts.md to pick a layout (the top has the Pre-flight class-name list and theme rhythm planning)
  6. When tuning details, read components.md to look up components
  7. After generating, read checklist.md to self-check (the top P0-0 rule enforces the pre-check)

Core design principles (philosophy)

These principles are distilled from 5 rounds of iteration on the "one-person company" talk deck. Violating any one of them makes the visual feel collapse.

  1. Restraint over flash: the WebGL background only shows through on hero pages, barely visible on ordinary pages
  2. Structure over decoration: no shadows, no floating cards, no padding boxes; all information relies on large type + typeface contrast + grid whitespace
  3. Content hierarchy is defined jointly by type size and typeface: largest serif = main title, medium serif = subtitle, large sans = lead, small sans = body, mono = metadata
  4. Images are first-class citizens: images are cropped only at the bottom, keeping the top and sides intact; grids use a fixed height:Nvh, not aspect-ratio stretching
  5. Rhythm relies on hero pages: hero and non-hero alternate, so the eyes don't tire
  6. Consistent terminology: Skills is just Skills, don't mix Chinese-English translation

Reference works

This skill's visual tone references:

  • Guizang's "One-Person Company: Organizations Folded by AI" talk (2026-04-22, 27 pages)
  • The layout of Monocle magazine
  • The demo from YC president Garry Tan's blog post "Thin Harness, Fat Skills"

You can treat these as style anchors.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Magazine Web Ppt AI skill do?

For marketing and gtm work: bind launches, campaigns, events, and brand plans to growth and pipeline outcomes. Built around the core query "annual-marketing-plan", with GTM strategy lead judgment, buyer-ready proof, and this outcome: approve launch plan, campaign budget, or GTM motion.

Why use Magazine Web Ppt on TypingMind?

Because you install it once and use it with any model. Magazine Web Ppt is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Magazine Web Ppt in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/nexu-io/open-design/tree/main/design-templates/guizang-ppt. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Magazine Web Ppt?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Magazine Web Ppt?

As many as you like. As long as a model supports skills, you can use Magazine Web Ppt with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Magazine Web Ppt AI skill free?

Yes. It is published on GitHub by nexu-io under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇