用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/mikailustuner/OmniRule --skill data-visualization命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
Bun runtime: HTTP server, file I/O, SQLite, test runner, package manager, bundler — all-in-one JS toolchain.
Clerk: Drop-in auth UI, Organizations, User management, JWT templates, webhooks, Next.js middleware integration.
Gelişmiş masaüstü, tarayıcı ve işletim sistemi kontrol yeteneği. Görsel (koordinat tabanlı) fare/klavye otomasyonu, DOM manipülasyonu, pencere yönetimi, gelişmiş dosya, ağ ve süreç yönetimini kapsar.
基于 SOC 职业分类
正在显示 SKILL.md
| name | data-visualization |
| description | Chart types, library selection, responsive design, and accessibility patterns for data visualization. |
| triggers | {"extensions":[".tsx",".ts"],"keywords":["chart","graph","d3","recharts","nivo","echarts","visualization","plot","dashboard"]} |
| auto_load_when | Building charts or data visualizations |
| agent | frontend-ops |
| tools | ["Read","Write","Bash"] |
Which chart to use?
Comparison:
├── Bar chart: Categorical comparisons, simple
├── Column chart: Time-based comparisons
├── Grouped/stacked: Multi-variable comparison
└── Bullet chart: Target vs actual
Distribution:
├── Histogram: Binned frequency
├── Box plot: Quartiles, outliers
├── Density plot: Continuous distribution
└── Violin plot: Distribution shape
Composition:
├── Pie/donut: Parts of whole (<6 segments)
├── Stacked area: Over time, trend emphasis
├── Treemap: Hierarchical composition
└── Sankey: Flow between states
Relationship:
├── Scatter: Two continuous variables
├── Bubble: Three continuous variables
├── Line: Continuous over time (trend)
└── Heatmap: Matrix relationships
Part-to-whole:
├── Donut: Simple, limited categories
├── Treemap: Hierarchical, many items
└── Sunburst: Multi-level drill-down
Library decision tree:
Small/simple (< 10kb):
├── Chart.js: Quick, canvas-based
├── ApexCharts: Good defaults, animations
└── Recharts: React-only, SVG-based
Medium/features:
├── ECharts: Powerful, many chart types
├── Nivo: React, highly customizable
└── Victory: React, declarative
Enterprise/advanced:
├── D3.js: Maximum control, steep learning
├── Highcharts: Commercial, excellent support
└── Plotly: Scientific, Python/JS
Framework-specific:
├── React: Recharts, Nivo, Visx
├── Vue: Vue-chartjs, ECharts wrapper
└── Angular: ngx-charts, Highcharts
Responsive strategy:
├── Aspect ratio: Maintain proportions
├── Mobile breakdown:
│ ├── Simplify: Fewer data points
│ ├── Scroll/zoom: Interaction patterns
│ ├── Touch: Larger touch targets
│ └── Labels: Abbreviate or hide
├── Breakpoints: 480px, 768px, 1024px
└── Debounce resize: Performance
Visual accessibility:
├── Color: Don't rely on color alone
│ ├── Add patterns/textures
│ ├── Use labeled legend
│ └── Provide data table alternative
├── Contrast: WCAG 4.5:1 minimum
├── Labels: Axis labels, data labels
└── Font size: 12px minimum
Screen reader:
├── ARIA labels on charts
├── Data table fallback
├── Role="img" with description
└── Announce updates via aria-live
Interactions to consider:
├── Tooltip: Hover details
├── Click: Drill-down, selection
├── Zoom/pan: Large datasets
├── Filter: Cross-filtering
├── Export: PNG/SVG download
└── Legend toggle: Show/hide series
Large datasets (>1000 points):
├── Sampling: Show subset, aggregate
├── Canvas: Better than SVG for many points
├── Virtualization: Only render visible
├── WebGL: For 3D or massive datasets
└── Aggregation: Server-side rollups
When to animate:
├── Initial load: Context setting
├── Update: Data changes, transitions
└── Highlight: Focus attention
When to avoid:
├── Real-time streaming: Choppy
├── Accessibility: Motion sensitivity
└── Print: No animation support
Consistency checklist:
├── Colors: Use design system palette
├── Typography: Match body/headings
├── Spacing: Consistent padding/margins
├── Grid: Align to 8px baseline
└── Components: Reusable chart wrapper
❌ Rendering 100,000 SVG elements directly in DOM
✅ Canvas or WebGL for large datasets; virtualize SVG lists
❌ Pie charts for comparing more than 4 values
✅ Bar chart for comparison; pie only for part-of-whole with <4 segments
❌ Dual Y-axis charts (misleading scale)
✅ Two separate charts or normalized data
❌ No loading state during data fetch
✅ Skeleton/placeholder chart while data loads
❌ Color as the only differentiator (accessibility)
✅ Color + pattern/shape; check contrast with colorblind simulation
| Chart type | When to use | Library |
|---|---|---|
| Bar | Comparison | Recharts / Chart.js |
| Line | Trend over time | Recharts / D3 |
| Scatter | Correlation | D3 / Observable Plot |
| Heatmap | 2D density | D3 |
| Treemap | Hierarchical part-of-whole | D3 |
| Large data | WebGL rendering | deck.gl / regl |