Skip to main content

component-design

Apply a coherent visual system, responsive behavior, interaction states, and accessibility to Kun components.

Datos de origen

Repositorio
KunAgent/Kun
Última actividad en el origen
30 de agosto de 2026 a las 03:07
Idioma detectado de SKILL.md
inglés
Estrellas
6372
Forks
593

Opciones de instalación

De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.

Revisa los archivos de origen

Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.

Explorador de archivos
3 archivos

Mostrando SKILL.md

SKILL.md
Instrucciones de origen · Vista previa de solo lectura
name
component-design
description
Apply a coherent visual system, responsive behavior, interaction states, and accessibility to Kun components.
# Component Design Copyright (c) 2026 KunAgent. Licensed under the MIT License. ## Purpose Apply a coherent visual system, responsive behavior, interaction states, and accessibility to Kun components. ## Tool routing | Tool or skill | Use | |---|---| | `design_system` | Create or apply project design tokens. | | `design_validate` | Validate design-system usage. | | `design_component` | Publish the component prototype. | | `generate_image` | Create a needed local visual asset. | ## Workflow 1. Define hierarchy, content density, primary action, and states. 2. Choose type, spacing, color, radius, and motion tokens before styling details. 3. Design mobile and desktop layouts intentionally. 4. Implement hover, focus, active, disabled, loading, empty, success, and error states as relevant. 5. Validate contrast, semantics, motion, and responsiveness. ## Completion gates - Use one focal point and avoid nested-card clutter. - Use color as reinforcement, not the only signal. - Keep remote resource policy least-privilege. ## Boundaries - Do not depend on undocumented host CSS or runtime globals. - Do not hide essential text in imagery or animation. ## Delivery Lead with the outcome, name the evidence used for verification, and disclose any real limitation that remains.
Ver en GitHub