Material Design 3 Ui logo

Material Design 3 Ui

Community
skydashnet
material-design-3-ui

Design, redesign, review, or implement user interfaces using Google Material Design 3 (M3), Material You, and optional Material 3 Expressive principles. Use for UI/UX screens, app flows, design systems, component selection, theming, responsive/adaptive layouts, accessibility audits, Figma-ready specifications, or developer handoff where Material Design 3 is required.

Overview

Publisherskydashnet
Repositorymaterial-design-3-ui-skill
Skill namematerial-design-3-ui
Stars
156
Forks
12
Bundled files
27
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.

  • 27 bundled files

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

  • Open source

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

Installation

Install the Material Design 3 Ui 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/skydashnet/material-design-3-ui-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/skydashnet/material-design-3-ui-skill.git \
  .claude/skills/material-design-3-ui
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Material Design 3 Ui 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 Material Design 3 Ui 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 Material Design 3 Ui 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.

Material Design 3 UI/UX

Purpose

Create interfaces that behave like Material Design 3, not interfaces that merely look rounded or “Google-like.”

Treat M3 as a system:

user goal → information architecture → hierarchy → adaptive layout → semantic tokens → components → states → interaction → motion → accessibility → visual expression

M3 Expressive is optional. Use it to strengthen hierarchy, usability, recognition, and emotional clarity; never use it as permission to make every element loud.

Source of truth

  1. Prefer current official Material 3 guidance at https://m3.material.io/.
  2. For Android implementation details, prefer https://developer.android.com/.
  3. Treat design guidance separately from implementation-library/API status.
  4. Verify current official guidance before asserting exact dimensions, dependency versions, experimental API names, or recently changed behavior.
  5. Do not silently mix Material 2 rules into Material 3.
  6. Pixel and Google-app screenshots are examples, not the specification.

Core rules

MUST

  • Start from the user’s task, content, hierarchy, and platform before styling.
  • Use semantic design tokens instead of scattered one-off values.
  • Choose components by purpose and behavior, not visual resemblance.
  • Prefer established M3/platform components when they solve the interaction.
  • Define important states: default, pressed, focus, hover where applicable, selected, disabled, loading, error, empty, and success where relevant.
  • Design for the available window rather than a guessed device category.
  • Preserve accessible contrast, target size, focus, semantics, scaling, and non-color state communication.
  • Respect system bars, safe areas, cutouts, keyboards/IME, foldable hinges, and edge-to-edge behavior where applicable.
  • Keep primary actions distinguishable from secondary and destructive actions.
  • Make custom components inherit the token, state, motion, and accessibility logic of the system.
  • Explain deliberate departures from M3 when product requirements justify them.

MUST NOT

  • Do not “Materialize” a screen by only adding large radii, pastel colors, shadows, or Material Symbols.
  • Do not hardcode raw colors throughout screens when semantic roles can be used.
  • Do not use color alone to communicate meaning or state.
  • Do not wrap every section or list row in a card.
  • Do not maximize corner radius on every component.
  • Do not use FABs for minor, destructive, or ambiguous actions.
  • Do not use chips as generic buttons or primary navigation.
  • Do not use tabs for unrelated top-level destinations.
  • Do not make all components equally expressive.
  • Do not invent a new control when an established M3/platform control already solves the interaction.
  • Do not sacrifice accessibility to preserve a visual composition.
  • Do not stretch a compact/mobile layout unchanged across wide windows.

Progressive-disclosure references

Load only the references needed for the current task. Do not read every reference by default.

NeedRead
Color roles, surfaces, dark theme, dynamic colorreferences/color-system.md
Type hierarchy, scaling, expressive typographyreferences/typography.md
Shape, cards, containment, elevationreferences/shape-and-elevation.md
Spacing, alignment, readable widths, safe areasreferences/spacing-and-layout.md
Buttons, FAB, cards, chips, selection, inputs, feedbackreferences/component-selection.md
Navigation bar/rail/drawer, tabs, app barsreferences/navigation.md
Forms, validation, settings, text inputreferences/forms-and-input.md
Snackbar, dialog, sheet, loading, empty/error statesreferences/feedback-and-overlays.md
Responsive, tablet, foldable, desktop, multi-panereferences/adaptive-design.md
Accessibility audit or any final UI reviewreferences/accessibility.md
Transitions, feedback animation, shape morphingreferences/motion.md
Material 3 Expressivereferences/m3-expressive.md
Final audit / suspicious “Material-looking” UIreferences/anti-patterns.md

For a broad end-to-end design, load the references incrementally as decisions require them. Always include the accessibility reference before final approval.

Workflow

Follow this order unless the task explicitly scopes one stage.

1. Understand the product

Identify the platform, primary user goal and action, top-level destinations, content hierarchy, data density, input methods, brand constraints, required states and edge cases, target window sizes, and whether M3 Expressive is appropriate.

If missing information does not block the work, make a conservative M3-aligned assumption and state it. Ask only when the missing information materially changes architecture or interaction.

2. Establish information architecture

Before styling:

  • Group related information and identify the primary task per screen.
  • Separate navigation from actions and distinguish destructive actions.
  • Remove duplicated controls and define progressive disclosure.
  • Avoid showing information merely because space is available.

3. Choose adaptive structure

Read references/adaptive-design.md and references/navigation.md when multiple window sizes or navigation forms matter.

Prefer canonical patterns such as list-detail, supporting pane, feed, and adaptive navigation when they fit the content.

Wider space must improve context or productivity rather than merely stretch content.

4. Build the theme semantically

Read the relevant color, typography, shape/elevation, and spacing references.

Use the hierarchy:

reference/system tokens → semantic/system roles → component tokens

Screens should depend on semantic roles rather than scattered literal values.

5. Select components

Read references/component-selection.md plus specialized navigation/form/feedback references as needed.

For each important control, determine its semantic purpose, emphasis, immediate or transactional behavior, interaction states, accessibility behavior, and adaptive behavior.

6. Define states and feedback

Cover relevant loading, empty, error, success, disabled, selected, pressed, focus, hover, and busy/submitting states.

Error states must provide a recovery path.

7. Apply accessibility

Read references/accessibility.md.

Accessibility is a release requirement. Do not defer it to visual polish.

8. Add motion

Read references/motion.md only when motion is part of the task.

Motion must explain state, hierarchy, spatial relationship, or response. Do not animate for spectacle.

9. Apply M3 Expressive if appropriate

Read references/m3-expressive.md.

Use a few deliberate expressive moments. Routine reading, forms, settings, and dense productivity surfaces should remain calm unless stronger expression improves usability.

10. Audit before approval

Read references/anti-patterns.md and run the self-audit below.

Self-audit

Before delivering a design, score each applicable category 0, 1, or 2.

  • 0 = incorrect / missing
  • 1 = partially correct / needs refinement
  • 2 = ready
CategoryCheck
Task clarityPrimary user goal and action are obvious
Information hierarchyGrouping and emphasis are coherent
Component semanticsControls match their actual behavior
Token disciplineSemantic roles are used consistently
Adaptive behaviorLayout improves across relevant windows
States & feedbackImportant states and recovery are covered
AccessibilityTargets, contrast, semantics, focus, scaling, reduced motion
Expressive restraintExpression improves hierarchy without creating noise

A design with any 0 in component semantics, states & feedback, or accessibility is not ready for approval.

Do not inflate scores to satisfy the user. State the concrete issue and correction.

Handoff contract

When asked to design, redesign, audit, or hand off a UI, provide enough detail for another designer or developer to reproduce the decisions.

Unless the user requests another format, include as applicable:

  1. Design intent — user goal and hierarchy.
  2. Layout — regions, panes, navigation, adaptive behavior.
  3. Theme roles — semantic colors, typography, shape, elevation.
  4. Component map — exact M3 component type/variant for important controls.
  5. States — interaction, loading, empty, error, success, disabled.
  6. Accessibility — target size, contrast, names, focus, keyboard, scaling.
  7. Motion — only meaningful transitions/state changes.
  8. Implementation notes — platform-specific caveats and experimental/unsupported APIs.
  9. Self-audit — include when the user asks for a review, audit, or production-readiness check.

When creating actual code or an artifact, apply these decisions instead of stopping at a description.

Semantic handoff example

Prefer:

text
Screen background: surface
Primary text: onSurface
Secondary text: onSurfaceVariant
Primary CTA: filled button / primary + onPrimary
Secondary CTA: outlined button
Section container: surfaceContainer
Subtle separator: outlineVariant
Error container: errorContainer + onErrorContainer

Avoid handoff based on arbitrary literals unless they come from the project’s token system.

Platform implementation notes

Android / Jetpack Compose

  • Prefer androidx.compose.material3 for new M3 work.
  • Prefer stable APIs for production unless alpha/experimental dependencies are explicitly accepted.
  • Do not assume every M3 Expressive API is stable.
  • Use Material theme roles rather than raw colors.
  • Follow current Android adaptive guidance for resizable and large-window experiences.
  • Treat edge-to-edge and system insets as layout concerns.
  • For Wear OS, use Wear Material 3 rather than mixing mobile components.

Do not hardcode library version numbers into generated implementation unless they are verified against current official release notes.

Web / other platforms

Material 3 guidance can inform the design even when an official implementation library does not expose every component.

  • reproduce semantics and token relationships, not Android-specific quirks,
  • preserve native platform accessibility and input conventions,
  • mark custom implementations clearly,
  • do not claim official platform availability without verification.

Final principle

Material Design 3 is a semantic, adaptive, accessible design system.

A successful M3 interface should remain coherent when brand colors change, the window resizes, dark theme turns on, text scales up, keyboard replaces touch, or expressive styling is reduced.

If the design only works because every surface is rounded and colorful, it is not a robust Material 3 design.

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 Material Design 3 Ui AI skill do?

Design, redesign, review, or implement user interfaces using Google Material Design 3 (M3), Material You, and optional Material 3 Expressive principles. Use for UI/UX screens, app flows, design systems, component selection, theming, responsive/adaptive layouts, accessibility audits, Figma-ready specifications, or developer handoff where Material Design 3 is required.

Why use Material Design 3 Ui on TypingMind?

Because you install it once and use it with any model. Material Design 3 Ui 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 Material Design 3 Ui in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/skydashnet/material-design-3-ui-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 Material Design 3 Ui?

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 Material Design 3 Ui?

As many as you like. As long as a model supports skills, you can use Material Design 3 Ui with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Material Design 3 Ui AI skill free?

Yes. It is published on GitHub by skydashnet 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 👇