Skill 详情
webdesign-review
Conducts comprehensive web design reviews and audits, not primary design creation.
使用前先检查
自动化审核只检查相关性,不代表安全审查或推荐。使用前请阅读来源中的说明。
SKILL.md
这段内容是审核时保存的快照。外部来源才是完整且最新的版本。
--- name: webdesign-review description: Meta-skill for comprehensive web design reviews. Orchestrates all 20 domain skills for systematic analysis of websites, designs, and prototypes against enterprise design standards. --- # Webdesign-Review > Systematic design reviews through orchestrated application of all domain skills. ## Scope Use this skill when: You are conducting a comprehensive design review of a website, prototype, or design. This meta-skill coordinates the analysis across all relevant design domains and ensures that no dimension is overlooked. ## Principles ### 1. Rigor Over Gut Feeling A design review follows a defined structure, not personal preferences. Every assessment must be substantiated by a specific rule or principle from the domain skills. ### 2. Prioritize by Impact Not all issues carry equal weight. Prioritize: accessibility and usability before aesthetics. Functional defects before stylistic preferences. ### 3. Constructive Criticism With Actionable Solutions Every identified weakness is accompanied by a concrete improvement recommendation. "This is bad" is not a review -- "This violates rule X; a better approach would be Y" is a review. ### 4. Completeness Through Checklists The checklists from individual domain skills form the foundation of the review. No item is skipped. ## Rules ### Review Workflow - **DO**: ALWAYS begin with an overall assessment before drilling into details. - **DO**: Work through the domains in the following order (by priority): 1. `accessibility` -- Accessibility (legal obligation, BFSG) 2. `usability` -- Usability 3. `ux-design` -- User experience and flows 4. `ui-design` -- Layout, hierarchy, components 5. `web-typography` -- Typography, readability, scale 6. `color-theory` -- Color, contrast, scheme 7. `responsive-design` -- Behavior across breakpoints 8. `navigation-design` -- Navigation structure and clarity 9. `images-media` -- Images, video, performance 10. `branding-identity` -- Brand consistency 11. `landing-pages` -- Conversion optimization (if landing page) 12. `customer-journey` -- User journey and touchpoints 13. `component-patterns` -- Component architecture 14. `ui-patterns` -- Interaction patterns 15. `visual-direction` -- Visual direction and trends 16. `design-trends` -- Currency and contemporaneity 17. `design-process` -- Process quality 18. `ai-design-workflow` -- AI integration in design 19. `website-audit` -- Technical quality 20. `agent-ui-design` -- Agent interfaces (if applicable) - **DO**: Walk through the corresponding checklist for each domain. - **DO**: Categorize findings by severity: Critical / Important / Recommendation. - **DON'T**: Skip domains just because they are "probably fine." - **DON'T**: Record subjective opinions without a rule reference as findings. ### Assessment Structure - **DO**: Structure every finding as: Domain > Rule > Finding > Severity > Solution. - **DO**: Attach screenshots or code examples as evidence where possible. - **DO**: Explicitly call out positive aspects -- not just deficiencies. - **DON'T**: Exceed 20 findings per domain. Prioritize the most impactful ones. ### Severity Levels - **Critical**: Accessibility violations (BFSG), functional defects, security issues. MUST be resolved immediately. - **Important**: Usability problems, inconsistencies, performance deficiencies. SHOULD be resolved before launch. - **Recommendation**: Stylistic improvements, optimizations, best practices. CAN be addressed in future iterations. ## Patterns ### Quick Review (30 Minutes) Focus on the top 6 domains: accessibility, usability, ux-design, ui-design, web-typography, color-theory. Checklist pass without deep analysis. ### Standard Review (2 Hours) All 20 domains. Checklists completed in full. Findings documented with severity and solution. Summary with top 5 priorities. ### Deep Audit (1 Day) Full review plus: screen reader testing, Lighthouse analysis, performance measur在 GitHub 阅读完整来源 (打开外部页面)