uxcel-lab/product-skills

ux-dashboard

Design or critique an analytics or data dashboard — which metrics earn a place, how to organize and prioritize them, and which chart fits each one.

Ver código fuente

Cómo usar

  1. Copia el comando y revisa el paquete antes de instalarlo en tu entorno de agente.
  2. Consulta las herramientas, credenciales y entornos necesarios en el documento original.
  3. Describe las entradas y el resultado deseado; contrasta la salida con las fuentes.
npx skills add https://github.com/uxcel-lab/product-skills --skill ux-dashboard

Antes de empezar

  • La instalación no incluye suscripción al modelo, cuota API ni permisos sobre tus archivos.
  • Las estrellas e instalaciones reflejan actividad, no una prueba independiente de calidad o seguridad.
Documento original del Skill

Contenido del repositorio de origen con títulos, ejemplos, código, tablas, enlaces e imágenes preservados.

Dashboard Skill

How this skill behaves (read first)

This is a generative skill, and dashboards have a signature over-design trap: cramming every available metric onto one screen. As the lessons put it — "when everything is important, nothing is." The default instinct is to show all the data; a good dashboard shows the few things that answer the user's question. So this skill gates:

  1. Establish who's looking and what decision they're making — that determines what belongs on the screen.
  2. Apply the always-true core — organization, hierarchy, and chart-fit principles that hold for any dashboard.
  3. Decide density, detail, and chart types deliberately — surface the context-dependent choices instead of defaulting to "show it all."

Then hand off to the audits — especially aesthetics (clutter/hierarchy is the make-or-break here).


Step 0 — Establish context before designing

  • Audience & the decision they make — when and why they look at this, and what action the data should drive. (User research is the real answer; ask if unknown.) This is the filter for what belongs.
  • Dashboard type:
  • Operational — real-time, monitoring, needs an immediate response (patient vitals, flight traffic). Comprehensive at a glance, frequent updates.
  • Analytical — for investigation and thoughtful decisions (a sales dashboard); not time-sensitive, supports drill-down.
  • Roles — one audience or many? Multiple roles → customization matters.
  • Platform — screen size shapes how many widgets fit and the responsive layout.

State assumptions if proceeding without answers.


The always-apply core (true for any dashboard)

  • Start from the user's question, not the data you have. For each widget ask: "what decision does this support?" If nothing, cut it. Let goals + user needs + data nature pick the widgets.
  • Don't overwhelm. Relevant few over comprehensive many. Density without hierarchy is a mess.
  • Prioritize for scanning. People scan, not read. Put the most important metric where the eye lands first; arrange by importance.
  • Group related content. Use containers, dividers, and white space so related metrics read as a group (common-region principle); section labels/titles act as visual anchors — make them distinct.
  • Use white space to segment and highlight — it improves legibility in data-dense screens and gives key numbers breathing room.
  • Keep structure consistent — shared alignment, title sizes, and margins across containers; this also makes responsive layouts scale.
  • Keep comparison data visible together. Don't bury things users need to compare behind tabs/pop-ups; never hide always-needed info. Use tabs only for content that isn't compared side by side.
  • Match the chart to the data (see reference below) — the single highest-leverage correctness decision.

The context-dependent decisions (surface, don't auto-apply)

DecisionApply whenAvoid / adapt whenDefault recommendation
Operational vs. analytical framingReal-time monitoring needs glanceable, comprehensive layoutAnalytical use wants drill-down, less time pressureMatch layout/refresh to the type from Step 0
Which metrics/widgets to showThe metric supports a real decision"Because we have the data" / vanity metricsRuthlessly cut to what answers the user's question
Level of detail / drill-downAnalytical dashboards benefit from overview → detailOperational dashboards needing everything visible at onceOverview first, progressive drill-down; don't tab-hide comparisons
Chart type per metricPick by data intent (trend/compare/part-of-whole/correlation)Decorative chart choices; pie charts for many categoriesMatch to data (reference below)
Customization by roleMulti-role systems (sales mgr vs. analyst)Single, well-known audience (adds needless complexity)Offer when roles genuinely differ
Real-time updatesOperational monitoringAnalytical review (constant refresh is noise)Refresh cadence matched to decision speed

Chart selection & honesty (quick reference)

Match the chart to the data intent:

  • Bar — compare discrete categories. Line — trends over time. Area — trend + volume over time. Pie/Donut — parts of a whole (few slices only). Scatter — correlation between two variables. Bubble — three dimensions. Radar — compare multiple variables (few). Candlestick — financial price ranges.

Chart honesty & clarity (non-negotiable — these mislead or exclude users):

  • Start axes at zero. A truncated baseline exaggerates differences and is misleading. (Deliberately misleading viz → flag with the dark-patterns audit.)
  • No 3D chart styling — it distorts perceived values.
  • Don't encode meaning with color alone — pair with icons/labels/shapes (→ accessibility).
  • Horizontal labels, readable annotations, consistent color/markers, appropriate intervals (don't crowd a year into daily ticks), standard sharp-cornered bars, focus/unfocus states for multi-line charts.

Validate the result (orchestration)

Hand-offs name each lens by its installable skill name. Invoke one only if that skill is installed; if it isn't, this skill's own core already carries these rules — proceed without it rather than blocking.

After generating or revising, hand the result to the audit lenses rather than declaring it done. These are candidate lenses — posture is set by docs/orchestration-policy.md, or route the whole thing through ux-design-review. Here, `heuristics`, `dark-patterns`, and `microcopy` are Tier A (auto-run); `accessibility` and `aesthetics` are Tier B (offered) — under an existing design system aesthetics is suppressed and accessibility narrows to usage. If the user invoked this skill for one specific thing, respect that scope.

  • `ux-aesthetics-audit` (Tier B — offer; suppress under a design system) — the dashboard make-or-break: clutter, visual hierarchy, white space, container grouping, "is one thing clearly most important?"
  • `ux-heuristics-audit` (Tier A) — recognition, visibility of status (esp. operational), user control (customization), keeping comparisons visible.
  • `ux-accessibility-audit` (Tier B — offer; narrow under a design system) — chart color not the sole encoding, contrast, labelled sections, keyboard access to interactive widgets.
  • `ux-dark-patterns-audit` (Tier A) — misleading visualization: truncated axes, 3D distortion, cherry-picked intervals.
  • `ux-microcopy-audit` (Tier A) — section titles, labels, and annotations are clear and consistent.

If aesthetics flags overload, return to the gate: cut widgets back to the ones that answer the user's question.


Common do/don't patterns

❌ Don't✅ Do
Show every metric you haveShow the few that support a decision
Equal visual weight for everythingPrioritize; most important where the eye lands first
Loose widgets floating on one planeGroup with containers, dividers, white space, labelled sections
Hide comparison data behind tabsKeep side-by-side comparisons visible together
Decorative chart choice / pie with 12 slicesMatch chart to data intent
Truncated y-axis / 3D barsZero baseline, flat charts — don't mislead
Color-only series legendColor + icon/label/shape
Vertical/diagonal axis labelsHorizontal, readable labels
Ship without checkingHand off to aesthetics + heuristics + accessibility

Source lessons (Uxcel)

del mismo repositorio

Más Skills

Todos los Skills
uxcel-lab
Comunidad

ux-tables

Design or critique data tables and lists — scannable structure, alignment rules (left for text, right for numbers), sorting/filtering, row interactions, and the right list type for the content. Trigger when the user asks to design or review a data table, data grid, admin table, list view, email/feed list, or asks how to present tabular or structured data.

Instalaciones / 1h
1
GitHub Stars
34
Actualizado
2 jul
uxcel-lab
Comunidad

pm-analytics

Run a product-analytics analysis that produces a decision, not a dashboard — start from the question, pick actionable over vanity metrics, separate signal from noise, segment (averages lie), pair quant with qual, and end in an insight→impact→action story. Catches the defaults Claude misses: vanity metrics, correlation-as-causation, reacting to small samples/short windows, survivorship/selection bias, unsegmented averages. Trigger when asked to analyze product data/metrics, plan what to measure,

Instalaciones / 1h
1
GitHub Stars
27
Actualizado
2 jul
uxcel-lab
Comunidad

pm-gtm-plan

Build a go-to-market plan that's a stage-appropriate system — customers, channels, pricing, timing, and partnerships that fit the company's maturity — not a generic launch checklist. Gates on GTM readiness (validated demand) and company stage, picks the right channel and entry timing, and treats the buyer journey as non-linear. Catches the defaults Claude misses: GTM-as-one-time-launch, stage-skipping, channel/timing blindness, funnel-forcing, partnerships-as-shortcut. Trigger when asked to crea

Instalaciones / 1h
1
GitHub Stars
27
Actualizado
2 jul
uxcel-lab
Comunidad

pm-product-review

Orchestrated review of a PM artifact (spec, PRD, OKRs, roadmap, prioritization, strategy). Routes by artifact type to the right audits, auto-runs the always-relevant assumption check, and returns one prioritized report. Trigger when the user asks for a full review of a PM document, "review my spec/roadmap/OKRs", "is this PRD solid", or general PM feedback — but NOT when they name a single specific audit.

Instalaciones / 1h
1
GitHub Stars
27
Actualizado
2 jul