30
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill mcouthon/agents --skill design- SKILL.md49.8 KB
Overview
This skill provides a complete playbook for premium UI design and implementation tailored to dashboards, SaaS UIs, admin interfaces, and any product that needs a polished, professional look. It covers design direction, core craft principles, and a 9-phase implementation roadmap that you can follow from typography to final polish. Use it to choose a personality, establish a consistent system, and avoid common pitfalls that create technical debt.
How this skill works
The skill guides you through three parts: choose a design direction (personality, color foundation, layout, and typography), apply core craft rules (4px grid, padding symmetry, depth strategy, color semantics), and execute a 9-phase implementation (typography, color, shadows, animation, components, layout, domain features, data display, pages). Each step includes concrete deliverables, CSS/Tailwind patterns, and implementation notes so designers and engineers can align quickly. Triggers include phrases like 'use design mode', 'design system', and 'design system upgrade' to enter full-access design guidance.
When to use it
- Starting a new dashboard, admin panel, or SaaS product and you need a coherent design system
- Upgrading an existing UI to be more polished, consistent, and accessible
- Defining visual personality and practical tokens before engineering work begins
- Building component libraries, design tokens, or Tailwind integrations
- Preparing product for dark mode, high-density data screens, or polished marketing screenshots
Best practices
- Commit to one design personality (or a deliberate blend) before styling components
- Use a 4px base grid and symmetrical padding to ensure visual rhythm and consistency
- Define HSL semantic tokens and integrate them with Tailwind using <alpha-value> for flexible opacity
- Choose one depth strategy (borders-only, subtle shadow, layered) and apply it consistently
- Treat controls as crafted components—avoid native form elements when you need full styling control
- Reserve color for meaning; use grayscale for structure and monospace for numeric data
Example use cases
- Designing a dense analytics dashboard that prioritizes scanability and chart clarity
- Reworking an enterprise admin with improved navigation, consistent cards, and elevation rules
- Building a design system repo with typography tokens, HSL color variables, and Tailwind config
- Adding reliable dark mode using border-based separation and adjusted semantic colors
- Creating data-focused components (tables, KPIs, chart palettes) with tabular numbers and mono fonts
FAQ
Yes—blend intentionally and map each choice to concrete implementation values (radius, spacing, shadows) so the mix stays coherent.
Should I always avoid native form elements?
Avoid native elements when you need full visual control; for simple internal tools where native accessibility is critical, weigh trade-offs and wrap native inputs with consistent styling.