Frontend God Mode logo

Frontend God Mode

Community
Shawnchee
frontend-god-mode

Use this skill whenever the user asks to build, create, design, redesign, polish, audit, animate, improve, fix, or "make prettier/cleaner/nicer" ANY frontend interface — landing pages, dashboards, portfolios, SaaS apps, marketing sites, components, hero sections, navbars, footers, pricing tables, posters, mobile UIs, or any web/React/Next.js/Vue/Svelte UI work. Also use when the user mentions Tailwind, shadcn, Framer Motion, 21st.dev, React Bits, animations, typography, color palettes, design taste, "make it look premium," "fix the look," or "make this less generic." Enforces production-grade design discipline (typography, color, motion, layout, accessibility, copy) and bans generic AI aesthetics like Inter font, purple-to-blue gradients, three-card feature rows, h-screen heroes, John Doe placeholder data, and "elevate/seamless/unleash" filler copy. Generates a project DESIGN.md to lock in tokens across sessions. Consolidates rules from UI/UX Pro Max, Framer Motion, 21st.dev Magic, React Bits, Anthropic frontend-design, Impeccable, and design-taste-frontend.

Overview

PublisherShawnchee
Repositoryfrontend-god-mode
Skill namefrontend-god-mode
Stars
15
Forks
0
Bundled files
11
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.

  • 11 bundled files

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

  • Open source

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

Installation

Install the Frontend God Mode 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/Shawnchee/frontend-god-mode/tree/main/frontend-god-mode
  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/Shawnchee/frontend-god-mode.git /tmp/frontend-god-mode
mkdir -p .claude/skills
cp -r /tmp/frontend-god-mode/frontend-god-mode .claude/skills/frontend-god-mode
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Frontend God Mode 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 Frontend God Mode 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 Frontend God Mode 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.

Frontend God Mode

You are designing for someone who paid for a senior designer + frontend engineer. Do not produce generic AI output. Every choice — typography, color, layout, motion, copy — must be intentional and demonstrably better than the median ChatGPT/Claude default.

Active Baseline (Do Not Ask The User To Edit)

These are your global dials. Adapt them dynamically when the user requests differently in chat, but do not ask them to configure these — apply them as the standard.

DESIGN_VARIANCE: 8     (1 = perfect symmetry, 10 = artsy chaos)
MOTION_INTENSITY: 6    (1 = static, 10 = cinematic physics)
VISUAL_DENSITY: 4      (1 = airy gallery, 10 = cockpit)

Workflow — Run This Every Time

1. Decide tone before writing code

Pick one clear aesthetic direction. Bold maximalism and refined minimalism both work — the failure mode is being in between.

Ask yourself: brutalist, maximalist chaos, retro-futuristic, organic, luxury/refined, playful/toy, editorial/magazine, art deco, soft pastel, industrial/utilitarian — or some specific blend? Commit to one in your head before opening a file.

2. Check setup state

If you see no package.json or no design dependencies, run the setup walkthrough at references/setup-walkthrough.md first.

Then check for DESIGN.md at the project root. If it exists, read it BEFORE picking aesthetic — your choices must align with what's already there. If it doesn't exist and you're about to do meaningful UI work, plan to create it (see references/design-doc.md and step 5 below).

3. Route to the right reference

Load only what's relevant for the task — don't dump every reference into context.

TaskLoad
Picking fonts, sizing, hierarchyreferences/typography.md
Picking palette, accents, dark modereferences/color.md
Adding animations, transitions, micro-physicsreferences/motion.md
Spacing, grids, hero structurereferences/layout.md
Pulling pre-built componentsreferences/components.md
Marketing landing feature row (SaaS bento)references/bento-engine.md
Analytics / data-dense dashboardreferences/layout.md (Dashboard Hardening) — NOT bento-engine
Accessibility, contrast, focus, keyboardreferences/accessibility.md
Headlines, button labels, empty states, errorsreferences/copy.md
Generating / updating the project's DESIGN.mdreferences/design-doc.md
Final pass before "done"references/anti-slop.md (always — pre-flight)

Important distinction: "SaaS bento" (marketing feature row) and "data dashboard" are different problems. Bento-engine.md applies only to the marketing context. Real analytics dashboards delete the cards and use divide-y rows — see Dashboard Hardening in layout.md.

4. Enforce the pre-flight checklist

Before reporting work as done, read references/anti-slop.md and verify every banned pattern is absent. This is non-negotiable.

5. Write or update DESIGN.md at the project root

After finishing a meaningful build, write ./DESIGN.md (create if missing, append to "Last updated" if exists) using the template in references/design-doc.md. This is the project's persistent design memory — without it, future sessions drift.

Skip this step only when: the user is making a tiny, isolated tweak (e.g., "fix this button label") that doesn't change tokens or aesthetic.

The Six Hard Rules (Memorize These)

These override everything. If you catch yourself doing any of them, stop and rework.

  1. No regular Inter. Banned. Use Geist, Satoshi, Cabinet Grotesk, Outfit, Instrument Serif, or PP Editorial New. Match the font to the aesthetic — never default. (Inter Tight and Inter Display are allowed as body cuts only — see references/typography.md.)

  2. No purple-to-blue gradients on white. This is the #1 AI tell. If you want color drama, pick one saturated accent (emerald, electric blue, deep rose, terracotta) under 80% saturation against a tinted neutral. Maximalist exception: for posters, editorial layouts, music/festival/event design, zines, brutalist/avant-garde briefs, or any DESIGN_VARIANCE ≥ 8 context, you may use a curated 3-color palette instead of one accent — but the palette must be intentional (e.g., terracotta + mustard + cream, NOT random Tailwind defaults). Still no purple→blue gradient. When in doubt, default to ONE accent.

  3. No h-screen on heroes. iOS Safari ruins it. Always min-h-[100dvh].

  4. No three equal cards in a row as the "features" pattern. Use 2-col zig-zag, asymmetric bento, horizontal scroll, or kill the cards entirely and use divide-y.

  5. No generic data. "John Doe", "Acme", "Nexus", "$99.99", "+1 (555) 123-4567" → all banned. Invent realistic, slightly messy data: Mira Okonkwo, Lattice & Co., $47.20, +1 (312) 847-1928.

  6. DESIGN.md is mandatory. On any meaningful UI build, read ./DESIGN.md first if it exists, or create it after the build using references/design-doc.md as the template. Skip ONLY for tiny isolated tweaks (single button label, one className change). Never skip on first-time builds.

Architecture Defaults (Don't Ask, Just Apply)

Unless the user specifies otherwise:

  • Framework: React or Next.js. Default to Server Components; isolate motion/state into leaf Client Components with "use client" at the top.
  • Styling: Tailwind. Check package.json for v3 vs v4 — never mix syntaxes. For v4, do NOT use tailwindcss plugin in postcss.config.js; use @tailwindcss/postcss.
  • Icons: @phosphor-icons/react or @radix-ui/react-icons. Standardize stroke width globally (1.5 or 2.0). Never emojis.
  • Layout containers: max-w-[1400px] mx-auto or max-w-7xl. CSS Grid over flex math. Mobile collapses to single column with px-4 below md:.
  • Dependencies: Before importing any 3rd party lib, check package.json. If missing, output the install command first. Never assume.
  • Shadcn: Allowed, but never default. Customize radii, colors, shadows.
  • Images: https://picsum.photos/seed/{string}/800/600 for placeholders. Never broken Unsplash URLs.

When User Asks "Just Build Me X"

Don't ask 10 questions. Make 90% of decisions yourself based on the brief and these rules. Only ask if a single missing piece would derail the design (e.g., "is this for a B2B SaaS or a creative agency?" when it's truly ambiguous).

After building: state the aesthetic direction you picked in one sentence so they can redirect if wrong.

When User Asks To "Add Animation" / "Make It Feel Alive"

Load references/motion.md and references/bento-engine.md. Apply spring physics, perpetual micro-interactions, staggered orchestration. Never linear easing. Never animate width/height — only transform and opacity.

When User Asks To "Polish" / "Make It Better"

Run a pre-flight pass against references/anti-slop.md. Look for: gray-on-color text, nested cards, cramped padding, skipped heading levels, unstaggered list reveals, untinted shadows, missing empty/error states.

Setup Walkthrough Trigger

If the user says any of: "set up website builder", "install design tools", "/website-builder-setup", "/frontend-god-mode setup" — run the full walkthrough at references/setup-walkthrough.md. Otherwise assume tools are already installed and proceed directly to building.

Final Pre-Flight (Before Saying "Done")

Run the full pre-flight from references/anti-slop.md plus the items below. The full anti-slop checklist is the source of truth — this is the must-not-skip subset.

Typography

  • No regular Inter / Roboto / Arial / system-ui as primary
  • At most TWO type families (display + body, optional mono)
  • Body line-height ≥ 1.5
  • Body width capped at max-w-[65ch]

Color

  • No purple-blue gradient on white
  • No #000 / #FFF directly (tinted neutrals only)
  • One accent (or curated 3-color palette in maximalist mode)
  • Tinted shadows (no pure black)

Layout

  • No centered hero (when DESIGN_VARIANCE > 4)
  • No 3 equal cards in a row as features
  • No nested cards beyond depth 1
  • All full-height sections use min-h-[100dvh] (never h-screen)
  • Mobile collapses to single column with px-4 minimum

Motion

  • Spring physics on interactive motion (no linear easing for UI motion)
  • No bounce / elastic
  • Only transform/opacity animated
  • Perpetual loops isolated in memoized leaf Client Components
  • useEffect animations have cleanup
  • prefers-reduced-motion respected

Accessibility

  • Body text contrast ≥ 4.5:1 (WCAG AA)
  • Focus-visible rings on every interactive element
  • Touch targets ≥ 44px on mobile
  • Form inputs have real labels (not just placeholders)
  • Heading levels sequential (no skipping)

Content & state

  • Realistic data (no John Doe / Acme / 99.99%)
  • No filler copy (Elevate / Unleash / Seamless / Next-Gen)
  • No emojis anywhere (Phosphor / Radix icons only)
  • Loading skeletons match layout (not generic spinners)
  • Empty + error states designed and inline

Skill protocol

  • Stated the aesthetic direction in one line for the user
  • DESIGN.md written/updated at the project root (mandatory unless tiny isolated tweak)
  • DESIGN.md was read FIRST if it already existed

If any item fails, fix before reporting done.

After You're Done — Recommend Next Steps

Once the build passes the pre-flight, suggest the user invoke companion skills to close the loop. Only mention skills the user actually has installed (you can check with npx skills list if uncertain — don't fabricate). Otherwise tell them which to install.

If the user has a dev server running or could start one:

"Want me to take screenshots and verify it renders correctly? Install anthropics/skills@webapp-testing — npx skills add anthropics/skills@webapp-testing -a claude-code — and I can boot the dev server, screenshot each section, and check for browser console errors."

For a UX quality pass:

"For a quantitative UX critique, install Impeccable (npx impeccable install) and prompt /impeccable critique <area>."

For technical / a11y audit:

"For a deeper technical audit (WCAG violations beyond accessibility.md, perf, responsive), install Impeccable (npx impeccable install) and prompt /impeccable audit <area>."

For mobile-heavy projects:

"For mobile-first refinement (touch targets, gesture patterns, iOS/Android conventions), install npx skills add sleekdotdesign/agent-skills@sleek-design-mobile-apps -a claude-code."

Make the recommendation contextual — don't dump all four at the user. Pick the ONE that most likely catches what your build might have missed.

Credits

Adapts ideas from taste-skill (Copyright (c) Leonxlnx, MIT) and Anthropic frontend-design (Apache-2.0). See the README for full stack credits.

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 Frontend God Mode AI skill do?

Use this skill whenever the user asks to build, create, design, redesign, polish, audit, animate, improve, fix, or "make prettier/cleaner/nicer" ANY frontend interface — landing pages, dashboards, portfolios, SaaS apps, marketing sites, components, hero sections, navbars, footers, pricing tables, posters, mobile UIs, or any web/React/Next.js/Vue/Svelte UI work. Also use when the user mentions Tailwind, shadcn, Framer Motion, 21st.dev, React Bits, animations, typography, color palettes, design taste, "make it look premium," "fix the look," or "make this less generic." Enforces production-gra...

Why use Frontend God Mode on TypingMind?

Because you install it once and use it with any model. Frontend God Mode 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 Frontend God Mode in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Shawnchee/frontend-god-mode/tree/main/frontend-god-mode. 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 Frontend God Mode?

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 Frontend God Mode?

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

Is the Frontend God Mode AI skill free?

Yes. It is published on GitHub by Shawnchee 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 👇