- Home
- Skills
- UX & UI Design
- Anti AI-Slop UI — Make Generated UIs Look Designed
Works with the AI Agents you already use
Anti AI-Slop UI — Make Generated UIs Look Designed
Audit and redesign generic AI-generated interfaces into distinct, high-craft visual directions.
Free
Anti AI-Slop UI — Make Generated UIs Look Designed
Example session with this skill installed
My landing page looks like a generic v0 template. It has the usual purple gradients, three cards, and 'Unlock your potential' copy. Make it look like a high-end Swiss dev tool.
- Read your context and instructions
- Compiled the anti ai-slop ui
I have audited the landing page code. The initial UI contained 9 AI tells, including default indigo gradients, centered hero text, and generic marketing fluff. I have refactored the interface following the Swiss/Utility direction.
AI-tell score: 9 → 1
Direction: Swiss/Utility — strict 12-column grid, Inter Tight + JetBrains Mono, ink + signal orange.
Fixed
- Typography: Replaced centered hero with left-aligned Inter Tight (3x body size). Added JetBrains Mono for technical labels.
- Colour: Removed purple gradients. Used #0A0A0A for ink and #FF5C00 for signal orange accents.
- Layout: Replaced the 3-card row with a 2-column feature breakdown and a technical spec table.
- Copy: Replaced "Unlock your potential" with "Deploy microservices in < 200ms."
Remaining (intentional): Rounded corners (8px) kept for button consistency with the existing component library.
Revised UI Code (Tailwind/React)
import React from 'react';
const LandingPage = () => {
return (
<div className="bg-[#FCFCFC] text-[#0A0A0A] font-sans selection:bg-[#FF5C00] selection:text-white">
{/* Header */}
<nav className="border-b border-[#EEEEEE] px-6 py-4 flex justify-between items-center font-mono text-sm uppercase tracking-wider">
<span>System_Log</span>
<div className="space-x-8">
<a href="#docs" className="hover:text-[#FF5C00]">Documentation</a>
<a href="#deploy" className="bg-[#0A0A0A] text-white px-4 py-2 hover:bg-[#FF5C00] transition-colors">Deploy</a>
</div>
</nav>
{/* Hero */}
<section className="max-w-7xl mx-auto px-6 pt-32 pb-20 grid grid-cols-12 gap-8">
<div className="col-span-8">
<h1 className="text-8xl font-bold tracking-tighter leading-[0.9] mb-8">
DEPLOY <br />
INFRA <br />
INSTANTLY.
</h1>
<p className="text-xl max-w-md border-l-2 border-[#FF5C00] pl-6 py-2">
Provision production-ready clusters in under 200ms. No cold starts. No configuration drift.
</p>
</div>
</section>
{/* Features Table - Breaking the 3-card slop */}
<section className="max-w-7xl mx-auto px-6 py-20 border-t border-[#EEEEEE]">
<table className="w-full text-left font-mono text-sm border-collapse">
<thead>
<tr className="uppercase text-[#888888] border-b border-[#EEEEEE]">
<th className="pb-4 font-normal">Metric</th>
<th className="pb-4 font-normal">Our Edge</th>
<th className="pb-4 font-normal">Status</th>
</tr>
</thead>
<tbody className="tabular-nums">
<tr className="border-b border-[#EEEEEE]">
<td className="py-6">Latency</td>
<td className="py-6 text-2xl text-[#FF5C00]">12ms avg</td>
<td className="py-6">Global-Ready</td>
</tr>
<tr className="border-b border-[#EEEEEE]">
<td className="py-6">Uptime</td>
<td className="py-6 text-2xl text-[#FF5C00]">99.999%</td>
<td className="py-6">Verified</td>
</tr>
</tbody>
</table>
</section>
</div>
);
};
export default LandingPage;
Next steps
- Review the JetBrains Mono font-weight for technical labels to ensure readability on low-DPI screens.
- Replace the placeholder "System_Log" with your actual brand SVG.
- Apply the
tabular-numsclass to any other data-heavy components in the app.
Connects securely to your tools. The creator never sees your data.
About this skill
AI-generated interfaces often suffer from a "samey" aesthetic, characterized by purple gradients, identical bento grids, and vague marketing copy. This skill audits frontend code to detect these generic patterns and replaces them with a deliberate, high-craft visual direction. It moves beyond the defaults of tools like v0 or generic LLM generations to produce professional, distinct user interfaces.
What it does
- Pattern detection identifies common "AI tells" like centered-everything heroes, meaningless stat tiles, and emoji icons.
- Visual direction mapping selects a specific design system (e.g., Swiss/Utility, Editorial, Technical Dark) based on brand goals.
- Typography overhaul replaces generic scales with intentional pairings and strict ratios.
- Color refinement bans default indigo-violet gradients in favor of 60-30-10 distribution and brand-tinted neutrals.
- Layout restructuring breaks repetitive card rows with asymmetric grids and varied component sizes.
- Copy cleaning strips out fluff words like "unlock" or "seamless" for concrete, action-oriented language.
How it works
- Audit the existing code or UI description against a known list of AI-generated patterns to establish a baseline score.
- Select one of six distinct visual directions to define the new design constraints.
- Refactor the UI components in priority order, starting with typography and color before moving to layout and motion.
- Re-audit the final output to ensure the "slop" score has been reduced to a minimum.
Frameworks & tools
This skill is framework-agnostic but excels with Tailwind CSS, React, and standard CSS. It utilizes design principles like concentric nesting, tabular numerals, and spring-based motion.
Why this beats prompting it yourself
Generic prompts usually result in the LLM's average training data, which is the source of the "AI slop" look. This skill enforces a specific audit-and-fix workflow that mandates design decisions, preventing the agent from falling back on safe, boring defaults.
Use cases
- Refactoring a landing page generated by v0 to look like a premium B2B SaaS product.
- Converting a generic dashboard into a high-density technical interface for developers.
- Cleaning up a mobile app UI to follow native platform conventions rather than web-templated looks.
Known limitations
Does not generate custom brand assets or logos. Requires existing UI code or a detailed layout description to perform the audit effectively.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for UX & UI Design.
Requires existing UI code or layout descriptions to audit. Does not generate custom illustrations or logos. Focuses on visual structure and copy, not backend logic.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
This skill has been used 10 times.
Trust & safety
Security scan
Security scan passed 5 days ago
- Free with an Agensi account
Permissions
No special permissions declared or detected