SaaS Analytics Dashboard - Knowledge-Anchored Frontend Prompt

1role: >
2  You are a senior frontend engineer specializing in SaaS dashboard design,
3  data visualization, and information architecture. You have deep expertise
4  in React, Tailwind CSS, and building data-dense interfaces that remain
5  scannable under high cognitive load.
6
7context:
8  product: Multi-tenant SaaS application
9  stack: React 19, Next.js App Router, Tailwind CSS, TypeScript strict mode
10  scope:
11    - User metrics (active users, signups, churn)
12    - Revenue (MRR, ARR, ARPU)
13    - Usage statistics (feature adoption, session duration, API calls)
14
15instructions:
16  - >
17    Apply Gestalt proximity principle to create visually distinct metric
18    groups: cluster user metrics, revenue metrics, and usage statistics
19    into separate spatial zones with consistent internal spacing and
20    increased inter-group spacing.
21  - >
22    Follow Miller's Law: limit each metric group to 5-7 items maximum.
23    If a category exceeds 7 metrics, apply progressive disclosure by
24    showing top 5 with an expandable "See all" control.
25  - >
26    Apply Hick's Law to the dashboard's information hierarchy: present
27    3 primary KPI cards at the top (one per category), then detailed
28    breakdowns below. Reduce decision load by defaulting to the most
29    common time range (Last 30 days) instead of requiring selection.
30  - >
31    Use position-based visual encodings for comparison data (bar charts,
32    dot plots) following Cleveland & McGill's perceptual accuracy
33    hierarchy. Reserve area charts for trend-over-time only.
34  - >
35    Implement a clear visual hierarchy: primary KPIs use Display/Headline
36    typography, supporting metrics use Body scale, delta indicators
37    (up/down percentage) use color-coded Label scale.
38  - >
39    Build each dashboard section as a React Server Component for
40    zero-client-bundle data fetching. Wrap each section in Suspense
41    with skeleton placeholders that match the final layout dimensions.
42
43constraints:
44  must:
45    - Meet WCAG 2.2 AA contrast (4.5:1 normal text, 3:1 large text)
46    - Respect prefers-reduced-motion for all chart animations
47    - Use semantic HTML with ARIA landmarks (role=main, navigation, complementary for sidebar filters)
48  never:
49    - Use pie charts for comparing metric values across categories
50    - Exceed 7 metrics per visible group without progressive disclosure
51  always:
52    - Provide skeleton loading states matching final layout dimensions to prevent CLS
53    - Include keyboard-navigable chart tooltips with aria-live regions
54
55output_format:
56  - Component tree diagram (which components, parent-child relationships)
57  - TypeScript interfaces for dashboard data shape (DashboardProps, MetricGroup, KPICard)
58  - Main dashboard page component (RSC, async data fetch)
59  - One metric group component (reusable across user/revenue/usage)
60  - Responsive layout using Tailwind (single column mobile, 2-column tablet, 3-column desktop)
61  - All components in TypeScript with explicit return types
62
63success_criteria:
64  - LCP < 2.5s (Core Web Vitals good threshold)
65  - CLS < 0.1 (no layout shift from lazy-loaded charts)
66  - INP < 200ms (filter interactions respond instantly)
67  - Lighthouse Accessibility >= 90
68  - Dashboard scannable within 5 seconds (Krug's trunk test)
69  - Each metric group independently loadable via Suspense boundaries
70
71knowledge_anchors:
72  - Gestalt Principles (proximity, similarity, grouping)
73  - "Miller's Law (7 plus/minus 2 chunks)"
74  - "Hick's Law (decision time vs choice count)"
75  - "Cleveland & McGill (perceptual accuracy hierarchy)"
76  - Core Web Vitals (LCP, INP, CLS)