Ui Style Extractor logo

Ui Style Extractor

Community
ivansong1981
ui-style-extractor

Extract visual design systems from existing websites into reusable markdown style guides, optional website demos, and optional app UI style guides. Use when analyzing a live site, screenshots, or harvested CSS to document colors, typography, spacing, layout, components, motion, and design rules, or when generating new website or app UI that should match a reference site's visual language without copying its content.

Overview

Publisherivansong1981
Repositoryui-style-extractor
Skill nameui-style-extractor
Stars
6
Forks
1
Bundled files
Instructions only
LicenseMIT
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by ivansong1981 on GitHub. Read the source before you install it.

Installation

Install the Ui Style Extractor 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/ivansong1981/ui-style-extractor/tree/main/.claude/skills/ui-style-extractor
  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/ivansong1981/ui-style-extractor.git /tmp/ui-style-extractor
mkdir -p .claude/skills
cp -r /tmp/ui-style-extractor/.claude/skills/ui-style-extractor .claude/skills/ui-style-extractor
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ui Style Extractor 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 Ui Style Extractor 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 Ui Style Extractor 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.

UI Style Extractor

Extract a site's visual language into a reusable Markdown guide, then use that guide to build original pages that still feel native to the source system.

Workflow

  1. Decide the output first.
    • Produce only a website style guide when the user wants documentation or prompt context.
    • Produce a website style guide plus demo page when the user wants website validation.
    • Produce a website style guide plus app UI style guide when the user wants the visual system translated into product UI.
    • Produce all three only when the user explicitly wants both website and app outputs.
  2. Build the evidence set before writing.
    • Prefer production CSS, design tokens, computed styles, and real DOM structure over screenshot-only guesses.
    • Inspect at least the homepage and one secondary page unless the user gives a tighter scope.
    • Use screenshots to capture composition, density, rhythm, imagery treatment, and overall mood.
  3. Duplicate the blank template and fill it section by section.
    • Use exact values where possible.
    • If a section is unsupported by evidence, write Not observed or Not present instead of inventing details.
  4. Distill repeated patterns into system rules.
    • Promote only repeated values into palette, spacing, radius, shadow, or motion tokens.
    • Keep one-off values in component notes unless they clearly define the brand language.
  5. Generate derivative outputs only after the website guide is stable.
    • Change the copy, product, and brand-specific content.
    • Preserve the design grammar: typography hierarchy, spacing rhythm, component treatment, and motion behavior.
    • For app guides, inherit the token layer first, then mark shell, workflow, and state patterns as adapted or inferred where necessary.
  6. Validate against the reference.
    • Compare screenshots side by side.
    • Tighten the guide if the generated page drifts from the source.

Extraction Rules

  • Prefer exact CSS values over visual estimation.
  • Treat screenshots as supporting evidence, not the primary source of tokens.
  • Record which pages were analyzed near the top of the guide.
  • Separate global rules from component-local exceptions.
  • Call out uncertainty explicitly when the source is ambiguous.
  • If the site has multiple modes or contrasting sections, document each mode rather than averaging them together.
  • Keep the final "Summary of Key Design Rules" strict enough that another model can follow it without re-reading the entire guide.

Demo Rules

  • Build original content in the extracted style; do not reproduce the reference site's copy wholesale.
  • Reuse the extracted tokens faithfully before adding custom flourishes.
  • Match the site's density and alignment logic, not just its colors and fonts.
  • Include only enough interactivity to demonstrate the system.
  • If fidelity depends on animation, gradients, or texture, implement them rather than omitting them.

App Guide Rules

  • Treat app UI guides as a derived artifact unless real product UI was also analyzed.
  • Explicitly separate inherited website tokens from adapted or inferred app patterns.
  • Define the product assumption near the top: dashboard, admin panel, consumer utility, and so on.
  • Keep app shells, data views, and forms consistent with the source system instead of defaulting to a generic SaaS aesthetic.
  • If the source is mostly marketing UI, reduce decorative gradients and oversized hero spacing in app contexts.

Output Conventions

  • Prefer <site>/<site>-ui-style.md for guide files when working in a multi-example repo.
  • Prefer <site>/demo.html for validation pages.
  • Prefer <site>/<site>-app-ui-style.md for app adaptation guides.
  • Keep reference captures close to the example, such as ref-viewport.png or ref-fullpage.png.
  • Keep guides in Markdown and keep component examples copyable.

Shared Resources

  • Use ../../../ui-style-extractor/assets/style-guide-template.md as the blank extraction template.
  • Use ../../../ui-style-extractor/assets/app-ui-style-template.md as the blank app adaptation template.
  • Read ../../../ui-style-extractor/references/extraction-workflow.md when you need the detailed page-selection and evidence-gathering workflow.
  • Read ../../../ui-style-extractor/references/app-adaptation-workflow.md when translating a website system into an app UI guide.
  • Read ../../../ui-style-extractor/references/quality-bar.md when deciding whether a guide or demo is complete enough to ship.

Install Note

  • This wrapper exists so Claude Code can discover the shared skill when this repository is open as a project.
  • For installation into another project or into ~/.claude/skills/, copy the canonical ui-style-extractor/ folder from this repository into that skills/ directory.

Frequently asked questions

What does the Ui Style Extractor AI skill do?

Extract visual design systems from existing websites into reusable markdown style guides, optional website demos, and optional app UI style guides. Use when analyzing a live site, screenshots, or harvested CSS to document colors, typography, spacing, layout, components, motion, and design rules, or when generating new website or app UI that should match a reference site's visual language without copying its content.

Why use Ui Style Extractor on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ivansong1981/ui-style-extractor/tree/main/.claude/skills/ui-style-extractor. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Ui Style Extractor?

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 Ui Style Extractor?

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

Is the Ui Style Extractor AI skill free?

Yes. It is published on GitHub by ivansong1981 under the MIT 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 👇