Ui Forge logo

Ui Forge

Community
dxkjuanjuan
ui-forge

5-stage UI design forge: Discover → Tokenize → Compose → Animate → Polish. Integrates 9 sub-skills (ui-ux-pro-max, awesome-design-md, ui-styling, motion-ui, frontend-design-direction, make-interfaces-feel-better, design-system, design, brand) + 9 reference sites (React Bits, Motion Primitives, Aceternity UI, Framer Motion, GSAP, Pageflow, Godly, Animate UI, GSAP Showcase) + Figma MCP. Covers landing pages, SaaS dashboards, mobile UI, component libraries, and advanced motion. Use when building any UI from scratch or from a Figma design.

Overview

Publisherdxkjuanjuan
Repositoryui-forge
Skill nameui-forge
Stars
3
Forks
0
Bundled files
Instructions only
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 dxkjuanjuan on GitHub. Read the source before you install it.

Installation

Install the Ui Forge 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/dxkjuanjuan/ui-forge/tree/main/skills/ui-forge
  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/dxkjuanjuan/ui-forge.git /tmp/ui-forge
mkdir -p .claude/skills
cp -r /tmp/ui-forge/skills/ui-forge .claude/skills/ui-forge
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ui Forge 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 Forge 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 Forge 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 Forge — 5-Stage Design Forge

A production-grade UI design workflow that forges pixel-perfect, accessible, animated interfaces from Figma designs or text descriptions.

When to Use

  • Building any UI from scratch: landing pages, dashboards, mobile apps, component libraries
  • Converting a Figma design to code
  • Creating branded interfaces that need design system consistency
  • Adding motion and micro-interactions to existing components
  • When the user says "build me a page", "design a UI", "implement this Figma", or "create a component"

Arguments

ArgDefaultDescription
--figma-url URL—Figma file URL (triggers Entry A)
--brand NAME—Brand from awesome-design-md (e.g. stripe, vercel, linear)
--skip PHASES—Comma-separated phases to skip (e.g. animate,polish)
--intensity LEVELautoMotion intensity: low/medium/high (auto-detected from product type)

The 5 Stages

Entry A (Figma) ──┐
                   ├─→ Phase 1: Discover → Phase 2: Tokenize → Phase 3: Compose → Phase 4: Animate → Phase 5: Polish → Delivery
Entry B (Text)  ──┘
                                                                                                              ↓
                                                                                                     [Figma writeback?]

Quality Gate Protocol

Each phase has a quality gate. If a gate fails:

  1. Roll back to the previous phase and adjust
  2. Maximum 2 retries per phase
  3. After 2 retries, report the issue and ask user for guidance

Phase 1: Discover — Direction Finding

Entry A: Figma Design

When --figma-url is provided:

  1. Use figma-developer-mcp → get_figma_data to extract node tree, components, styles
  2. Map Figma structures:
    • Auto Layout → Flexbox (reliable)
    • Nested Auto Layout → CSS Grid (reliable)
    • Auto Layout wrap → Flex wrap (reliable)
    • Absolute positioning + Constraints → Responsive (needs manual adjustment)
    • Figma Variants → Component variants (partial support)
  3. Extract component instances → identify reusable components
  4. Extract Figma Variables → direct CSS variable conversion

Entry B: Text / Brand Description

When no Figma URL:

  1. Use ui-ux-pro-max to search style direction:
    bash
    python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "<user description>" --domain style
    python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "<user description>" --domain product
  2. Use awesome-design-md to match brand tokens — read ~/.claude/skills/awesome-design-md/design-md/<brand>.md
  3. Use frontend-design-direction to set: purpose, audience, tone, memorable detail, constraints

Mixed Mode

When user provides "reference this site + change to dark style":

  • Extract layout/structure from the reference
  • Take colors/fonts from brand tokens
  • Merge both into the design direction

Reference Sites

SiteHow to use
GodlyBrowse similar product sites for visual direction. Use kimi-webbridge to navigate godly.website and filter by category.
PageflowFor narrative/scrolling pages, reference scrollytelling structures. Use kimi-webbridge to browse pageflow.app.

Accessibility Pre-constraint

  • Verify primary color contrast ratio ≥ 4.5:1 at this stage
  • Confirm accessibility target: AA or AAA

Quality Gate Output

yaml
design-direction.md:
  product_type: SaaS | landing | mobile | component-library
  style_direction: [ui-ux-pro-max style name + AI prompt keywords]
  brand_token_source: [awesome-design-md brand] | [custom]
  reference_sites: [specific pages from Godly/Pageflow]
  figma_node_id: [if applicable]
  tech_stack: [detected from project]
  interaction_mode: SPA | SSR | SSG | native
  accessibility_level: AA | AAA

Skip Condition

  • Project has DESIGN.md < 30 days old AND requirements unchanged → skip
  • DESIGN.md > 30 days old → suggest re-discovery

Phase 2: Tokenize — Design Token Generation

Token Generation Flow

  1. Brand token base → awesome-design-md brand → extract colors/typography/spacing/rounded/shadows
  2. Style enhancement → ui-ux-pro-max search:
    bash
    python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "<style>" --domain style
    python3 ~/.claude/sskills/ui-ux-pro-max/scripts/search.py "<style>" --domain typography
    python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "<style>" --domain color
    python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "<style>" --design-system --variance N --motion N --density N
  3. Token systemization → design-system skill → three-layer tokens (primitive → semantic → component)
  4. Figma variable mapping (if applicable) → Figma Variables → CSS variables

Reference Sites

SiteHow to use
Animate UIReference for motion timing/easing tokens
GSAPStandard motion durations: fast 150ms, normal 300ms, slow 600ms

Accessibility Pre-constraint

  • All color palettes validated for contrast at generation time
  • Dark mode tokens generated simultaneously, not deferred

Quality Gate Output

Generate design-tokens.json with:

css
/* Primitive tokens */
:root {
  /* Colors — light */
  --color-primary: oklch(68% 0.21 250);
  --color-canvas: oklch(98% 0 0);
  --color-ink: oklch(18% 0 0);
  /* Colors — dark */
  --color-primary-dark: oklch(72% 0.21 250);
  --color-canvas-dark: oklch(15% 0 0);
  --color-ink-dark: oklch(95% 0 0);
  /* Typography */
  --text-display-xl: clamp(3rem, 1rem + 7vw, 8rem);
  --text-body: clamp(1rem, 0.92rem + 0.4vw, 1.125rem);
  /* Spacing */
  --space-xs: 4px; --space-sm: 8px; --space-md: 16px;
  --space-lg: 32px; --space-xl: 64px;
  --space-section: clamp(4rem, 3rem + 5vw, 10rem);
  /* Rounded */
  --rounded-xs: 4px; --rounded-sm: 8px; --rounded-md: 12px;
  --rounded-lg: 16px; --rounded-xl: 24px; --rounded-full: 9999px;
  /* Shadows */
  --shadow-subtle: 0 1px 2px rgba(0,0,0,0.05);
  --shadow-medium: 0 4px 12px rgba(0,0,0,0.08);
  --shadow-prominent: 0 8px 24px rgba(0,0,0,0.12);
  /* Motion */
  --motion-fast: 150ms; --motion-normal: 300ms; --motion-slow: 600ms;
  --ease-default: cubic-bezier(0.16, 1, 0.3, 1);
  --ease-bounce: cubic-bezier(0.34, 1.56, 0.64, 1);
  --ease-spring: cubic-bezier(0.22, 1, 0.36, 1);
  --stagger-default: 50ms; --stagger-fast: 30ms; --stagger-slow: 80ms;
  --distance-micro: 4px; --distance-small: 8px;
  --distance-medium: 16px; --distance-large: 32px;
}

Skip Condition

  • Project has token system + consistent with Phase 1 direction + brand unchanged → skip
  • Brand or style changed → do not skip
  • Minor tweaks → skip, do incremental update

Phase 3: Compose — Component Building

Component Generation Order (Strict Priority)

  1. Layout skeleton → Layout / Header / Main / Footer / Sidebar
  2. Core blocks → Hero / Data tables / Primary interaction areas
  3. Reusable components → Button / Card / Badge / Input / Select
  4. Auxiliary components → Toast / Tooltip / Modal / Dropdown

Build skeleton first, then fill in. Never generate all components at once.

UI Type Routing

TypeStrategySub-skills
Landing pageHero/CTA/Feature/Testimonial structurebanner-design + ui-styling
SaaS dashboardTable/Form/Modal/Sidebar/Chartsui-styling + design-system
Mobile UIResponsive + platform patternsui-styling + responsive tokens
Component libraryshadcn/ui base + custom variantsui-styling
Advanced motion3D Canvas + particles + Lottie placeholdersui-styling + motion scaffolds

Component Reuse Detection

Before generating, scan existing project components:

  • Has shadcn Button? → Don't regenerate
  • Has custom Card? → Check token compatibility, reuse if compatible
  • Use find and grep to detect existing components

Accessibility Pre-constraint

  • Component skeletons include ARIA attributes from the start
  • Interactive components include keyboard navigation from the start
  • Touch targets ≥ 44px

Reference Sites

SiteHow to use
React BitsSmall components + micro-interaction snippets. Browse reactbits.dev for copy-ready patterns.
Aceternity UIHigh-end motion components (3D Card, Spotlight, Beam, etc.). Reference ui.aceternity.com.
shadcn/uiBase component specs. Reference ui.shadcn.com for accessible component patterns.

Quality Gate Output

yaml
component-tree.md:
  tech_stack: [detected]
  generation_order: [layout → core → reusable → auxiliary]
  component_tree:
    Layout → Header / Main / Footer
    Header → Nav / Logo / CTA
    Main → Hero / Features / CTA
  reusable_components: [list with token compatibility check]
  file_structure: src/components/{hero,features,ui}/
  responsive_breakpoints: 320 / 768 / 1024 / 1440
  code_skeletons: [props + structure + class names only, not full implementation]

Skip Condition

  • Existing component library compatible with token system → skip
  • Only needs animation/polish → skip

Phase 4: Animate — Motion Injection

Motion Intensity Auto-Detection

Product TypeIntensityTypical Motion
Landing pageHighScroll narrative, parallax, entrance, particles
SaaS dashboardLowMicro-interactions, state transitions, skeleton screens
Mobile UIMediumGestures, page transitions, haptic feedback
Component libraryMicrohover, focus, active, loading states

Override with --intensity high|medium|low.

Motion Pattern Matching

Motion TypeImplementationReference
Entrancemotion/react whileInView / GSAP ScrollTriggerMotion / Motion Primitives
State transitionAnimatePresence + motion variantsFramer Motion
Scroll narrativeGSAP Timeline + ScrollTriggerGSAP / Pageflow
Micro-interactionCSS transitions / motion/react whileHover/whileTapAnimate UI / React Bits
3D / ParticlesThree.js / ts-particles / LottieGSAP Showcase

Using ui-ux-pro-max for GSAP Patterns

bash
python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "scroll reveal" --domain gsap
python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "stagger entrance" --domain gsap
python3 ~/.claude/skills/ui-ux-pro-max/scripts/search.py "page transition" --domain gsap

Tech Stack Motion Mapping

StackMotion Library
React + Tailwindmotion/react + GSAP (for complex)
Next.js + Tailwindmotion/react + GSAP (for complex)
Vue + Nuxt@vueuse/motion
Sveltesvelte/motion + svelte/transition
SwiftUI.animation() + withAnimation
FlutterAnimationController + ImplicitlyAnimatedWidget

Accessibility Pre-constraint

  • ALL motion must support prefers-reduced-motion
  • Reduced motion fallback: opacity transitions only, duration: 0ms
  • Implement via:
css
@media (prefers-reduced-motion: reduce) {
  *, *::before, *::after {
    animation-duration: 0.01ms !important;
    transition-duration: 0.01ms !important;
  }
}
tsx
// motion/react
const shouldReduceMotion = useReducedMotion()

Reference Sites

SiteHow to use
Framer MotionReact motion standard library. Reference motion.dev for API patterns.
Motion PrimitivesBase animation modules (entrance, exit, spring). Reference motion-primitives.com.
GSAPProfessional animation (ScrollTrigger, Timeline, SVG path). Reference gsap.com.
GSAP ShowcaseTop-tier motion case studies. Reference gsap.com/showcase.
Animate UIPure CSS motion fallback (hover, scroll-triggered). Reference animate-ui.com.

Quality Gate Output

yaml
animation-spec.md:
  intensity: high | medium | low
  entrance: { type, duration, stagger, easing }
  scroll: { type, trigger, scrub }
  micro: { hover, focus, active }
  states: { loading, success, error, empty }
  reduced_motion: { fallback, duration }
  implementation: [motion code per component]

Skip Condition

  • --skip animate flag
  • SaaS dashboard + user explicitly declines motion
  • Simple utility page with no interaction

Phase 5: Polish — Micro-interaction Refinement

1. Visual Polish (make-interfaces-feel-better)

Concentric radius: outer radius = inner radius + padding Optical alignment: offset asymmetric icons by 1-2px Layered shadows: transparent, subtle, background-compatible Text wrapping:

  • text-wrap: balance on headings and short titles
  • text-wrap: pretty on body text, captions, descriptions
  • Neither on long prose or code Tabular numbers: font-variant-numeric: tabular-nums for counters, prices, timers Image outlines: outline: 1px solid rgba(0,0,0,0.1); outline-offset: -1px Font smoothing: -webkit-font-smoothing: antialiased on macOS

2. Interaction State Completion

Every interactive component must have:

  • hover — visual feedback on mouse-over
  • focus — visible focus ring (2px offset, brand color)
  • active — pressed/engaged state
  • disabled — reduced opacity + not-interactive
  • loading — skeleton or spinner
  • error — error message + red accent
  • empty — empty state illustration or message

Touch targets ≥ 44px. Keyboard navigation complete.

3. Brand Consistency Check (brand)

  • Token usage: all colors/fonts/spacing strictly follow tokens
  • Tone: copywriting style matches brand voice
  • Dark mode: all components have complete dark mode variants
  • No hardcoded values outside the token system

4. Responsive Verification

  • 320px / 768px / 1024px / 1440px breakpoints
  • No text overflow, no image clipping, no layout shift
  • clamp() for fluid typography and spacing
  • min() / max() for boundary constraints

5. Accessibility Final Verification

  • prefers-reduced-motion all motion degrades correctly
  • All text meets AA (4.5:1) or AAA (7:1) contrast
  • ARIA attributes complete on all interactive elements
  • Keyboard navigation reaches every interactive element
  • Screen reader order matches visual order

Reference Sites

SiteHow to use
React BitsMicro-interaction detail patterns. Reference reactbits.dev.

Quality Gate Output

yaml
polish-checklist.md:
  visual: [concentric-radius ✓ optical-align ✓ shadows ✓ text-wrap ✓ tabular-nums ✓ image-outlines ✓ font-smoothing ✓]
  interaction: [hover ✓ focus ✓ active ✓ disabled ✓ loading ✓ error ✓ empty ✓]
  brand: [token-consistency ✓ tone ✓ dark-mode ✓]
  responsive: [320 ✓ 768 ✓ 1024 ✓ 1440 ✓ no-overflow ✓]
  accessibility: [reduced-motion ✓ focus-ring ✓ contrast ✓ ARIA ✓ keyboard ✓]

Final Deliverables

src/components/**/*.tsx    — Component code
src/styles/tokens.css      — CSS variables
src/styles/tokens.json     — Token JSON (for Figma writeback)
DESIGN.md                  — Design direction document
polish-checklist.md        — Polish checklist

Figma Writeback (when Entry A was used)

  • New tokens → write back to Figma Variables
  • Component variants → update Figma component library
  • Forms Figma ↔ Code bidirectional loop

Skip Condition

  • --skip polish flag

Sub-Skill Call Map

PhaseSub-SkillsReference Sites
Discoverfrontend-design-direction, ui-ux-pro-max, awesome-design-md, figma-developer-mcpGodly, Pageflow
Tokenizedesign-system, ui-ux-pro-max, awesome-design-mdAnimate UI, GSAP
Composeui-styling, banner-design, design-systemReact Bits, Aceternity UI, shadcn/ui
Animatemotion-ui, ui-ux-pro-max(--domain gsap)Framer Motion, Motion Primitives, GSAP, GSAP Showcase, Animate UI
Polishmake-interfaces-feel-better, brandReact Bits

9 Reference Sites — Complete Assignment

#SitePhasesRole
1React Bits3, 5Component snippets + micro-interaction details
2Motion Primitives4Base animation modules
3Aceternity UI3High-end motion component reference
4Framer Motion4React motion standard library
5GSAP2, 4Token standards + professional animation
6Pageflow1, 4Narrative structure + scroll narrative
7Godly1Similar product visual direction
8Animate UI2, 4Motion tokens + CSS motion fallback
9GSAP Showcase4Top-tier motion case studies

Usage Examples

# Full pipeline from text
ui-forge "Build a Stripe-style pricing page for my SaaS"

# From Figma design
ui-forge --figma-url https://figma.com/design/abc123/My-Design

# Skip animation for a dashboard
ui-forge "Admin dashboard with user management" --skip animate

# Specific brand with high motion
ui-forge "Product launch landing page" --brand vercel --intensity high

# Component library only
ui-forge "Design system button variants" --skip animate,polish

Accessibility Throughout

Not just Phase 5 — every phase has pre-constraints:

PhaseConstraint
DiscoverPrimary contrast ≥ 4.5:1, confirm AA/AAA target
TokenizeAll palettes contrast-validated, dark mode generated simultaneously
ComposeSkeletons include ARIA + keyboard navigation
AnimateAll motion supports prefers-reduced-motion
PolishFinal verify: contrast + ARIA + focus + keyboard + reduced-motion

Tech Stack Adaptation

StackPhase 3 SourcePhase 4 Library
React + Tailwindshadcn/ui + Radixmotion/react
Next.js + Tailwindshadcn/ui + Radixmotion/react + GSAP
Vue + NuxtNuxt UI@vueuse/motion
Svelte + SvelteKitshadcn-sveltesvelte/motion
SwiftUINative components.animation()
FlutterMaterial/CupertinoFlutter animations

Frequently asked questions

What does the Ui Forge AI skill do?

5-stage UI design forge: Discover → Tokenize → Compose → Animate → Polish. Integrates 9 sub-skills (ui-ux-pro-max, awesome-design-md, ui-styling, motion-ui, frontend-design-direction, make-interfaces-feel-better, design-system, design, brand) + 9 reference sites (React Bits, Motion Primitives, Aceternity UI, Framer Motion, GSAP, Pageflow, Godly, Animate UI, GSAP Showcase) + Figma MCP. Covers landing pages, SaaS dashboards, mobile UI, component libraries, and advanced motion. Use when building any UI from scratch or from a Figma design.

Why use Ui Forge on TypingMind?

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

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

Which AI models can use Ui Forge?

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 Forge?

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

Is the Ui Forge AI skill free?

It is published on GitHub by dxkjuanjuan. 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 👇