Design Style logo

Design Style

Community
yanliudesign
design-style

Recommend and apply design styles from a curated library of 67 visual aesthetics. Use when the user wants to choose a design style for their project, get style recommendations, or apply a style's CSS/colors/fonts to their code.

Overview

Publisheryanliudesign
Repositorydesign-style-skill
Skill namedesign-style
Stars
15
Forks
3
Bundled files
69
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.

  • 69 bundled files

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

  • Open source

    Published by yanliudesign on GitHub. Read the source before you install it.

Installation

Install the Design Style 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.
    https://github.com/yanliudesign/design-style-skill/tree/main
  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/yanliudesign/design-style-skill.git \
  .claude/skills/design-style
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Design Style 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 Design Style 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 Design Style 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.

Design Style

Recommend and apply design styles from a curated library of 67 visual aesthetics — from Brutalism to Glassmorphism, Cyberpunk to Japandi.

When This Skill Activates

  • User asks "what style should I use for my project?"
  • User says "recommend a design style"
  • User mentions a specific style name ("use glassmorphism", "apply brutalist style")
  • User wants to apply colors, fonts, or CSS from a design aesthetic
  • User asks "帮我选设计风格" / "推荐设计风格" / "选个设计风格"

Core Principles

  1. Show, don't lecture — Lead with concrete CSS, colors, and prompt templates. Not theory.
  2. Context-aware — A fintech dashboard needs different styles than an indie game landing page.
  3. Opinionated — Recommend 3 styles, not 15. Explain why each fits.
  4. Immediately actionable — Every run ends with a browsable comparison HTML the user can pick from, plus paste-ready CSS variables.
  5. No middle stops — After the user provides purpose + basic info, run straight through to a generated preview. Don't stop to ask "which one do you prefer?" — let the visual do the choosing.

Phase 0: Detect Mode

  • Mode A: Recommend (default) — User wants styles picked for their project. Go to Phase 1.
  • Mode B: Lookup — User names a specific style ("tell me about glassmorphism"). Skip Phase 1/2, go straight to Phase 3 for that style.
  • Mode C: Apply — User picks a style and wants it applied to existing code ("apply cyberpunk to this file"). Skip Phase 1/2/3, go straight to Phase 4.

Phase 1: Two Questions — Purpose + Basic Info

Combine both into a single AskUserQuestion call. Do NOT ask them one at a time.

Question 1 — Purpose (目的)

Header: Purpose Question: "What are you building, and what's the goal? / 你在做什么?想达成什么目的?" Options (with freeform allowed):

  • SaaS / 工具产品
  • 作品集 / 个人站
  • 电商 / 品牌店
  • 博客 / 内容站
  • Landing Page / 营销落地页
  • App / 移动端
  • 演示 / Slides / 演讲
  • 其他 (freeform)

Question 2 — Basic Info (基本信息)

Header: Basics Question: "Tell me a bit more — brand vibe, audience, must-have/avoid colors, any constraints. / 说几句项目基本信息:品牌调性、目标用户、必须保留或规避的颜色、任何限制。" No fixed options — freeform text answer.

Example acceptable answers:

  • "AI dev tool, targeting engineers, must feel technical, no baby pastels"
  • "独立摄影师作品集,想要暗色高级感,偏冷色调"
  • "儿童教育 App,家长掏钱,得让家长放心又让孩子觉得有趣"

Skip Phase 1 entirely if the user's initial message already covers both (e.g., "帮我给我的 AI dev SaaS 找 3 个大胆但可信的风格" — that answers Purpose + Basics inline).


Phase 2: Pick 3 Styles + Auto-Generate Preview

No middle stop. After Phase 1, immediately do all of the following in one turn:

Step 2.1 — Read the reference

Read styles-reference.md — the "Use Case Recommendations" section for the user's Purpose category, and the full detail blocks (### N. id — Name / 中文名) for candidate styles.

Step 2.2 — Infer innovation level from Basic Info

Parse the free-text answer for signals:

Signal words in BasicsInnovation level
trustworthy, professional, safe, financial, enterprise, 稳重, 可信, 专业Level 1
clean, modern, standard SaaS, 现代, 干净Level 2
bold, distinctive, signature, memorable, 大胆, 有辨识度, 有记忆点Level 3
experimental, rule-breaking, avant-garde, portfolio-grade, 先锋, 实验, 冒险, 作品集级Level 4

If ambiguous, default to Level 2–3 (the sweet spot for most projects).

Step 2.3 — Choose 3 styles with real spread

Pick 3 styles that span the axis — not 3 flavors of the same thing. Rules of thumb:

  • One "safest" pick that clearly matches the level.
  • One "signature" pick that leans one level bolder.
  • One "wildcard" from an adjacent Purpose category that could reframe the project (e.g., for a SaaS pick, throw in an Editorial or Kinetic option).

Avoid recommending 3 styles that share the same DNA quadrant (e.g., three dark-neon styles). The 3 should feel visually distinct.

Step 2.4 — Print the recommendation card (3 blocks)

## Style Recommendations

### 1. [Name] · [中文名] — [one-line reason it fits this project]
Colors: ■ #hex ■ #hex ■ #hex ■ #hex
DNA: Roundness X | Contrast X | Warmth X
Why: [2 sentences tied to the user's Purpose + Basics — not generic style copy]

### 2. [Name] · [中文名] — [one-line reason]
...

### 3. [Name] · [中文名] — [one-line reason]
...

Step 2.5 — Generate a side-by-side comparison HTML

Immediately (in the same turn — do NOT ask the user "which one do you prefer?"), create a single-file HTML that renders the same sample content in all 3 chosen styles side-by-side:

  • Save path: ~/Desktop/Claude skills/design-style-preview-<project-slug>.html
  • Sample content: a hero card the user can imagine as their landing page — brand name (from Basics if given, else "Your Project"), a tagline, a primary CTA button, 3 feature chips. Same content in each of the 3 style renders.
  • Layout: 3 columns on desktop (grid-template-columns: repeat(3, 1fr)), stacked on mobile.
  • Each column displays: style number & name at top, then the rendered card using that style's CSS from styles-reference.md.
  • Include a "Pick this one" button under each render that copies the style ID to clipboard.
  • Open the file with open <path> at the end of the turn.

Step 2.6 — Print the follow-up prompt

End the turn with:

"Preview opened. Tell me which one clicks — or say 'wildcard 3 more' and I'll pull a different trio."


Phase 3: Show Style Details (Lookup Mode)

Only used when the user names a specific style directly (Mode B) or after they pick one from the Phase 2 preview.

Read the style's full entry in styles-reference.md and present:

  1. Overview — Name, difficulty, DNA scores
  2. Color Palette — All colors with their roles, as a visual table
  3. CSS Snippet — The core CSS that defines this style
  4. Font Recommendations — Extract from the CSS/prompts what fonts work best
  5. Prompt Templates — All 3 prompt templates (Basic, Advanced, Keywords)
  6. Do's and Don'ts — Quick reference for staying on-style
  7. Related Styles — 3 similar styles for exploration

Then offer: "Want me to apply this to your project? (Phase 4)"


Phase 4: Apply to Project

Triggered when the user picks a favorite from the Phase 2 preview OR explicitly asks to apply a style.

4.1 Generate CSS Variables

css
:root {
  /* [Style Name] Theme */
  --color-primary: #...;
  --color-secondary: #...;
  --color-accent: #...;
  --color-background: #...;
  --color-surface: #...;
  --color-text: #...;
  --color-text-secondary: #...;

  --font-heading: '...', ...;
  --font-body: '...', ...;

  --radius-sm: ...;
  --radius-md: ...;
  --radius-lg: ...;

  --shadow: ...;
}

4.2 Scan the user's project

Read existing CSS/HTML/framework files to detect: Tailwind, vanilla CSS variables, CSS Modules, styled-components, and any existing theme tokens.

4.3 Apply in the project's own idiom

  • Tailwind → theme.extend config with the palette, fonts, shadows.
  • CSS Variables → a :root block (or update existing tokens).
  • Styled-components / CSS Modules → a theme object.
  • Raw HTML → inject a <style> block or inline styles.

Also provide:

  • Google Fonts / Fontshare <link> for recommended fonts
  • A sample component (card, button, nav) styled in the chosen aesthetic
  • The 3 prompt templates for generating more UI with AI tools

4.4 Verify

  • WCAG AA contrast on primary/text combos
  • Font links properly included
  • No overrides that break existing components

Data Source

All style data comes from styles-reference.md:

  • Quick index of all 67 styles with colors, difficulty, and tags
  • Use case recommendations by project type and innovation level
  • Full details per style: colors, CSS, prompts, do's/don'ts, DNA scores

The authoritative interactive reference: https://design-lab-yanliu.vercel.app/


Guardrails

  • Never recommend more than 3 styles at once — too many options paralyze.
  • Never stop between Phase 1 and Phase 2's HTML output — the point of this skill is to let the user see the preview before they decide, not to interview them further.
  • Always ground recommendations in the user's Purpose + Basics — no generic "these are popular styles."
  • When applying styles, preserve existing code structure — don't rewrite files.
  • If the user's project has an established design system, integrate with it rather than override.
  • Font recommendations must use Google Fonts or Fontshare (free, no licensing issues).
  • Check color contrast when applying dark or low-contrast styles.

Bundled files

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

and 9 more files.

Frequently asked questions

What does the Design Style AI skill do?

Recommend and apply design styles from a curated library of 67 visual aesthetics. Use when the user wants to choose a design style for their project, get style recommendations, or apply a style's CSS/colors/fonts to their code.

Why use Design Style on TypingMind?

Because you install it once and use it with any model. Design Style 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 Design Style in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/yanliudesign/design-style-skill/tree/main. 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 Design Style?

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 Design Style?

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

Is the Design Style AI skill free?

It is published on GitHub by yanliudesign. Check the repository for licensing terms. 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 👇