frontend-design

Generate and audit production-grade frontend interfaces with distinctive aesthetics that avoid generic AI design patterns.

Updated Aug 2, 2026
One-click install
npx skills add https://github.com/leonardoacosta/skills --skill frontend-design-leonardoacosta
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: frontend-design
Source: https://github.com/leonardoacosta/skills/tree/main/leo-design/skills/frontend-design
Command: npx skills add https://github.com/leonardoacosta/skills --skill frontend-design-leonardoacosta

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? AI-generated web interfaces tend to converge on the same generic look — purple gradient heroes, Inter fonts, glassmorphism cards, and three-column feature grids. This Skill guides the creation of distinctive, production-grade frontend code and audits existing UI against a catalog of known AI design anti-patterns. ## Core Features & Use Cases - Create, Audit, and Spot-fix modes: Build new components, pages, and apps; audit existing UI against a 34-pattern anti-slop canon with severity rankings (STRUCTURAL/SURFACE/COSMETIC); or diagnose and rewrite a single flagged element. - Design system awareness: Reads a project-level DESIGN.md as the normative token source of truth, exports consumable Tailwind/DTCG tokens, and validates contrast against it. - Deep reference library: Ships references for OKLCH color craft, typography mechanics, execution-craft checks, 10 pre-built themes, a 66-brand DESIGN.md catalogue, icon sourcing (Phosphor, svgl, thesvg), dithering/ASCII effects, Motion/GSAP animation components, and 91 Radiant shader effects. - Use Case: Ask for a landing page for a fintech product — the Skill commits to a bold aesthetic direction, picks a distinctive font pairing, writes working HTML/CSS/React to docs/diagrams/, and avoids every named slop pattern. ## Quick Start Ask the agent to build a distinctive landing page or audit your existing UI against the anti-slop canon, and it will produce polished frontend code in docs/diagrams/.

Frequently Asked Questions about frontend-design

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I avoid generic AI-generated design in my frontend?

Commit to one bold aesthetic direction, pick a distinctive display and body font pairing, and check your output against the anti-slop canon — patterns like purple gradient heroes, glassmorphism cards, and three-column feature grids. The Skill names each pattern and prescribes a specific replacement.

How do I audit existing UI for AI design slop?

Share the UI code or copy and request an audit. The Skill checks the project's DESIGN.md tokens first, then runs the anti-slop canon and microcopy references, naming each violation and ranking findings as STRUCTURAL, SURFACE, or COSMETIC severity.

Does the Skill support project design systems like DESIGN.md?

Yes. When a DESIGN.md exists at the project root, it is the normative source of truth. The Skill parses its YAML tokens, exports them via npx @google/design.md to Tailwind or DTCG formats, and validates output contrast with the lint command.

Where does the Skill write generated HTML and CSS files?

All generated HTML, CSS, and standalone frontend files are written to docs/diagrams/ in the current working directory. It never writes to /tmp, the project root, or other locations.

Can I make my UI match a specific brand like Linear or Stripe?

Yes. The brand-design-catalogue reference provides 66 DESIGN.md files extracted from real brands. Fetch the target brand's file, apply its color, typography, spacing, and motion tokens, and follow the one-brand-per-page rule.

What icon libraries does the Skill recommend for UI and logos?

Phosphor is the default for generic UI chrome, svgl.app for dev and SaaS brand logos, Boxicons bxl-* as a brand fallback, and thesvg for cloud architecture icons. Emojis are never used as icon placeholders.