Frontend Design logo

Frontend Design

Community
Waishnav
frontend-design

Create distinctive, production-grade frontend interfaces with strong art direction. Use when building or restyling web components, pages, dashboards, landing pages, React apps, HTML/CSS layouts, or any UI where originality, polish, and visual quality are critical. Enforce anti-slop rules, section-level layout diversity, typography quality, and interaction design.

Overview

PublisherWaishnav
Repositoryself-improvement-frontend-design-skill-loop-for-codex
Skill namefrontend-design
Stars
13
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 Waishnav on GitHub. Read the source before you install it.

Installation

Install the Frontend Design 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/Waishnav/self-improvement-frontend-design-skill-loop-for-codex/tree/main/experiments/_archived/version-3/.agents/skills/frontend-design
  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/Waishnav/self-improvement-frontend-design-skill-loop-for-codex.git /tmp/self-improvement-frontend-design-skill-loop-for-codex
mkdir -p .claude/skills
cp -r /tmp/self-improvement-frontend-design-skill-loop-for-codex/experiments/_archived/version-3/.agents/skills/frontend-design .claude/skills/frontend-design
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Frontend Design 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 Design 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 Design 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.

Build real, production-grade frontend code with intentional visual direction. Counter common GPT/Codex failure modes: text-heavy layouts, repeated section architecture, generic style fallback, weak contrast, and shallow interactions. Prioritize user constraints over this skill: framework, libraries, coding style, architecture, and file conventions always win. This skill defines design outcomes and quality bars, not mandatory implementation tooling.

Portability and Conflict Resolution (Required)

  • Apply design guidance through the user's chosen stack and conventions.
  • Do not force specific tools or libraries unless the user asks for them.
  • Express decisions as transferable intent (hierarchy, rhythm, contrast, interaction feedback), then map to local primitives.
  • Implement tokens/themes using the project's native system (variables, utility config, theme objects, component tokens, etc.).
  • If preferences are missing, choose a minimal-dependency approach and keep it easy to migrate.
  • Reuse existing project patterns before introducing new abstractions.

Workflow (Required)

  1. Lock context quickly.
  • Identify audience, core task, tone, platform, and constraints.
  • If context is missing, make practical assumptions and proceed.
  1. Generate five distinct directions internally before coding.
  • Force real divergence in section architecture, composition, type system, and color strategy.
  • Pick one direction and commit.
  1. Define a direction contract before implementation.
  • Concept statement (one sharp phrase).
  • Font pair (display + body).
  • Token palette (dominant, support, accent, neutral).
  • Spatial rule (asymmetry, strict grid, split narrative, or layered canvas).
  • Signature element (one memorable motif).
  • Section graph id (from the Section Graph Library below).
  • Primary interaction mode (from Interaction Modes below).
  1. Implement a token system first.
  • Define reusable primitives for type scale, spacing, color, radius, borders, elevation, and timing in the project's preferred format.
  • Build components from those primitives, not ad hoc values.
  1. Implement section architecture with deliberate breaks.
  • Enforce the selected section graph.
  • Add at least two macro composition breaks after the hero.
  1. Add interaction depth.
  • Add at least one hero-level interaction moment.
  • Add at least two component-level interactions for pages, one for small components.
  1. Run quality gates and revise once.
  • Do not stop at first draft quality.

Hard Guards for Known Failure Modes

Text-heavy output:

  • Keep hero headline <= 12 words.
  • Keep hero supporting copy <= 24 words.
  • Keep above-the-fold body copy to two short blocks maximum.
  • Convert long explanations into cards, bullets, metrics, or visual modules.

Section architecture repetition:

  • Treat each concept as a unique section graph, not a palette variation.
  • Do not repeat the same section order across concepts.
  • If generating multiple concepts, no duplicate section graph ids.

Structural monotony:

  • Do not use the same hero composition more than once across concepts.
  • Do not run three consecutive sections with the same container style (for example: card-grid, card-grid, card-grid).
  • Add at least one full-bleed or off-grid section in each concept.

Readability regressions:

  • Reject low-contrast text and clipped typography.
  • Reject blur/noise effects that reduce legibility.
  • Reject decorative layers that obscure content.

Generic defaults:

  • Do not default to Inter/Roboto/Arial/system as primary identity unless explicitly requested.
  • Do not default to purple-pink-blue neon gradients.
  • Do not default to generic "SaaS hero + three cards + CTA" structure.
  • Do not use fake terminal aesthetics unless context explicitly justifies it.

Reliability regressions:

  • Do not ship broken scroll behavior.
  • Do not hide critical content behind fixed overlays.
  • Do not ship obvious horizontal overflow on mobile unless intentionally designed and controlled.

Section Graph Library (Use as Families, Not Templates)

Use one graph as a structural spine, then style it uniquely.

  • G1 Narrative Split: Hero split -> Capability rail -> Workflow strip -> Proof grid -> Pricing -> Footer.
  • G2 Editorial Flow: Manifesto hero -> Sticky chapter sections -> Model table -> Membership CTA -> Footer notes.
  • G3 Cockpit Stack: Command hero -> Status panels -> Inspector rail -> Queue/process timeline -> Pricing dock -> Footer.
  • G4 Gallery First: Media hero -> Collection mosaic -> Detail spotlight -> Compare slider -> CTA bar -> Footer.
  • G5 Poster Web: Typographic hero -> Framed statement blocks -> Feature ribbons -> Offer slab -> Footer tape.
  • G6 Neo Utility: Hard-edge hero -> Utility matrix -> Constraint/protocol section -> Feature ledger -> Pricing card wall.
  • G7 Quiet Luxury: Minimal hero -> Signature benefit band -> Refined cards -> Testimonial quote plane -> Pricing strip.
  • G8 Story Scroll: Hero scene -> Sticky storytelling sequence -> Key metrics board -> Offer reveal -> Footer.
  • G9 Swiss Data: Typographic masthead -> Data rows -> Comparison matrix -> Spec blocks -> CTA row -> Footer.
  • G10 Hybrid Canvas: Asymmetric hero -> Floating module cluster -> Tooling/process section -> Proof collage -> Pricing.

Macro Composition Breaks (Required)

Each concept must include at least two of these:

  • Dense to sparse transition (or sparse to dense).
  • Grid to freeform overlap transition.
  • Light to dark surface transition (or inverse).
  • Static section followed by motion-led section.
  • Text-led section followed by media-led section.
  • Orthogonal layout followed by diagonal or layered composition.

Interaction Modes

Select one primary mode per concept. For multi-concept outputs, do not reuse the same primary mode.

  • I1 Kinetic typography hero.
  • I2 Scroll scene transitions (sticky sections or progressive reveals).
  • I3 Interactive comparison surface (tabs, segmented control, or switchable views).
  • I4 Procedural background motion (subtle and readable).
  • I5 Cursor-responsive accent behavior (restrained, never distracting).
  • I6 Card choreography (stagger, parallax, or progressive detail reveal).
  • I7 Data-panel live state feel (status, counters, or activity rhythm).

Modern Pattern Atlas (Recent, Reusable)

Use these as pattern families, not copy-paste templates.

  • Bento product narrative: asymmetrical modular blocks with clear hierarchy.
  • Editorial mono-luxe: serif headlines, precise grid, restrained ornament.
  • Neo-brutalist utility: sharp borders, flat planes, heavy contrast, intentional rawness.
  • Quiet luxury SaaS: muted neutrals, premium spacing, subtle motion.
  • App cockpit: dense command surfaces, inspector rails, status modules.
  • Kinetic typography hero: oversized type as primary visual engine.
  • Story-scroll narrative: sticky scenes and section-driven transitions.
  • Gallery-first interface: media-dominant composition with lightweight chrome.
  • Soft-material depth: layered surfaces, controlled blur, tactile lighting.
  • Swiss data grid: strict rhythm, typographic rigor, restrained accents.
  • Experimental poster-web: expressive type, overlap, framing, and composition tension.

Composition Rules

  • Establish a clear first-view focal point.
  • Use one spatial thesis consistently.
  • Balance one dense zone with one breathing zone.
  • Use scale contrast, overlap, or framing to create hierarchy.
  • Design mobile intentionally; do not only collapse desktop.

Typography Rules

  • Pair one display face with one body face.
  • Keep font families to two maximum.
  • Use a fluid or stepped type scale appropriate to the stack and platform, with intentional line length.
  • Apply case, tracking, and weight as part of brand voice.
  • For five-concept outputs, assign explicit type pairs up front and do not reuse the exact same pair on more than two routes.
  • Give each route one visible typographic signature move (for example: drop cap intro, split italic phrase, vertical kicker, or condensed all-caps strapline).
  • Control hero line breaks intentionally so the first fold reads as designed rhythm, not browser-default wrapping.
  • Run a typography audit before final output; if two routes share nearly identical hierarchy, weight map, and rhythm, revise one route.

Color and Surface Rules

  • Tokenize palette in the project's native theming system (bg, fg, muted, accent, accent-2, surface, border semantics).
  • Keep one dominant chroma and sparse accents.
  • Validate text/control contrast.
  • Use gradients or textures only when conceptually justified.

Interaction and Motion Rules

  • Add one orchestrated entrance sequence.
  • Add meaningful hover/focus/active states.
  • Keep a coherent timing system (for example: 140ms, 220ms, 360ms).
  • Prefer native platform motion primitives first; add heavier JS animation only when state complexity requires it.
  • Respect reduced motion preferences.

Multi-Concept Protocol

When asked for multiple designs:

  • Generate all concepts in one run.
  • Default to five concepts unless the user specifies a different count.
  • Enforce unique fingerprints across concepts:
  • Section graph id
  • Hero composition type
  • Primary interaction mode
  • Typography family
  • Palette temperature
  • Section rhythm
  • If concepts are shown in one app, provide a lightweight, stack-appropriate concept switcher.

Delivery Checklist

  • Production-ready implementation that runs in the requested stack without manual fixes.
  • Domain-specific copy (no placeholder or vague generic slogans).
  • Visible focus styles and readable contrast.
  • Desktop and mobile both polished.
  • Scroll behavior verified end to end.
  • One post-build refinement pass completed.

Self-Critique Quality Gate (Score 0-2)

  • Distinctiveness
  • Section-graph uniqueness
  • Layout originality
  • Text economy
  • Typography craft
  • Interaction quality
  • Accessibility and contrast
  • Responsiveness
  • Scroll and overflow reliability

If any score is below 2, revise before final output.

Frequently asked questions

What does the Frontend Design AI skill do?

Create distinctive, production-grade frontend interfaces with strong art direction. Use when building or restyling web components, pages, dashboards, landing pages, React apps, HTML/CSS layouts, or any UI where originality, polish, and visual quality are critical. Enforce anti-slop rules, section-level layout diversity, typography quality, and interaction design.

Why use Frontend Design on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Waishnav/self-improvement-frontend-design-skill-loop-for-codex/tree/main/experiments/_archived/version-3/.agents/skills/frontend-design. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Frontend Design?

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 Design?

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

Is the Frontend Design AI skill free?

It is published on GitHub by Waishnav. 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 👇