- Home
- Skills
- Frontend & Web Apps
- Design Quality
Works with the AI Agents you already use
Design Quality
Full-cycle design quality agent to audit, unslop, and build high-craft user interfaces.
Free
Design Quality
Example session with this skill installed
Review this dashboard UI code. It feels a bit generic and the spacing is off. Does it look like AI slop?
- Read your context and instructions
- Compiled the design quality
I have analyzed the dashboard component. Three parallel subagents (Slop Auditor, Scorecard Scorer, and Emotion Analyst) have completed their review.
Slop Audit
- High Severity: Found generic purple-to-blue linear gradients on primary action buttons.
- Medium Severity: Card grid uses placeholder emoji icons instead of a consistent SVG library.
- Medium Severity: "Inter-everything" typography with no weight contrast, typical of default AI outputs.
Scorecard
Product fit: 5/10 (Functional but lacks specific industry character)
Hierarchy: 4/10 (Primary actions blend into the background navigation)
Clarity: 6/10 (Labels are clear, but affordances are weak)
Taste: 3/10 (Relies on glow effects and rounded corners rather than layout)
Restraint: 4/10 (Too many competing shadows and borders)
Density: 7/10 (Information volume is appropriate for a dashboard)
Craft: 4/10 (Non-concentric border radii and inconsistent 13px margins)
Responsiveness: 8/10 (Flexbox layout holds up well at mobile widths)
Overall: 5.1/10
Emotional Design
- Felt State: The interface feels "busy" yet "empty." The user lacks a clear focal point upon landing.
- Peak-End: The submission success state is a generic browser alert, missing a branded "Moment."
Verdict
Best part: Clean layout density and responsive behavior.
Worst part: Heavy reliance on visual slop (glows/gradients) to mask poor hierarchy.
Required fix (P0): Fix spacing to 8px grid; remove hue gradients from surfaces.
Quick wins (P1): Replace emoji with Lucide icons; set headings to text-wrap: balance.
Next steps
- Would you like me to dispatch the Fix Mode agents to align the spacing and color ramp?
- I can generate a
PROJECT_DESIGN.mdto define a unique character and signature for this dashboard. - Should I run a deep-dive on the Visual Hierarchy (scored 4/10)?
Connects securely to your tools. The creator never sees your data.
About this skill
AI-generated interfaces often suffer from "slop": generic layouts, neon gradients, and poor information density. This skill provides a specialized framework to audit, critique, and fix UI code by applying rigorous design principles and specialized subagent reviews.
What it does
- Slop detection identifies generic AI artifacts like purple gradients, emoji icons, and filler card grids.
- Scored critiques evaluates designs across 8 dimensions including Hierarchy, Taste, Density, and Craft.
- Automated deep-dives dispatches specialized subagents for low-scoring areas like typography or color.
- Fix passes executes a 10-step sequential refinement process covering spacing, accessibility, and motion.
- Direction setting defines product character and "Design Dials" before building new interfaces.
How it works
- Input resolution where the agent analyzes screenshots, HTML/CSS, or project folders to establish context.
- Mode selection triggers specific workflows for reviewing existing work, creating new UI, or fixing broken designs.
- Subagent dispatch uses parallel workers to audit slop, score the interface, and map emotional design arcs.
- Iterative refinement applies P0-P3 fixes and polishes micro-interactions like concentric radii and font smoothing.
Frameworks & tools
Works with React, Tailwind CSS, Vue, Svelte, and standard HTML/CSS. It integrates with icon libraries like Lucide or Heroicons and utilizes OKLCH color spaces for modern token systems.
Why this beats prompting it yourself
Generic prompts result in generic layouts that look like every other AI bootstrap template. This skill enforces an 8px grid, accessibility standards, and a "One Bright Element" rule to ensure interfaces feel intentional rather than generated.
Use cases
- Auditing a frontend codebase for design debt and accessibility failures.
- Unslopping a generic dashboard to make it look like a professional SaaS product.
- Establishing a consistent design direction for a new feature from scratch.
- Polishing micro-interactions and typography for a production-ready landing page.
Known limitations
Cannot confirm visual styles from text descriptions alone. Requires source code or images for scoring. Technical performance checks are limited to basic surface-level analysis.
Requires visual input or source code for accurate auditing. Cannot evaluate performance beyond basic CSS checks. Does not support non-standard proprietary UI frameworks.
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 any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 12 days ago
- Free with an Agensi account
Requires
Permissions
The skill reads source files to audit styles and components. No external API calls. Write Files is used only when applying fix passes to the codebase.