Back to Skills

design-shiny-ui

pjt222
Updated 2 days ago
6 views
17
2
17
View on GitHub
Metadesign

About

This skill helps developers build modern, responsive Shiny app UIs using bslib for theming and layout_columns for grids. It covers implementing value boxes, cards, custom CSS/SCSS, accessibility, and brand consistency. Use it when creating new apps, modernizing old fluidPage apps, or improving responsiveness and accessibility.

Quick Install

Claude Code

Recommended
Primary
npx skills add pjt222/agent-almanac -a claude-code
Plugin CommandAlternative
/plugin add https://github.com/pjt222/agent-almanac
Git CloneAlternative
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/design-shiny-ui

Copy and paste this command in Claude Code to install this skill

Documentation

Design Shiny UI

Responsive, accessible Shiny UI → bslib + modern layouts + custom CSS.

Use When

  • New Shiny UI from scratch
  • Modernize app: fluidPage → bslib
  • Brand theming (colors, fonts)
  • Responsive across screens
  • A11y improvements

In

  • Required: App purpose + audience
  • Required: Layout type (sidebar, navbar, fillable, dashboard)
  • Optional: Brand colors + fonts
  • Optional: Custom CSS/SCSS (default: bslib only)
  • Optional: A11y reqs (WCAG lvl)

Do

Step 1: Pick Page Layout

bslib page constructors:

# Sidebar layout — most common for data apps
ui <- page_sidebar(
  title = "My App",
  sidebar = sidebar("Controls here"),
  "Main content here"
)

# Navbar layout — for multi-page apps
ui <- page_navbar(
  title = "My App",
  nav_panel("Tab 1", "Content 1"),
  nav_panel("Tab 2", "Content 2"),
  nav_spacer(),
  nav_item(actionButton("help", "Help"))
)

# Fillable layout — content fills available space
ui <- page_fillable(
  card(
    full_screen = TRUE,
    plotOutput("plot")
  )
)

# Dashboard layout — grid of value boxes and cards
ui <- page_sidebar(
  title = "Dashboard",
  sidebar = sidebar(open = "closed", "Filters"),
  layout_columns(
    fill = FALSE,
    value_box("Revenue", "$1.2M", theme = "primary"),
    value_box("Users", "4,521", theme = "success"),
    value_box("Uptime", "99.9%", theme = "info")
  ),
  layout_columns(
    card(plotOutput("chart1")),
    card(plotOutput("chart2"))
  )
)

→ Layout matches nav + content needs.

If err: layout off → check using page_sidebar() / page_navbar() (bslib) not fluidPage() / navbarPage() (base shiny). bslib versions → better defaults + theming.

Step 2: Config bslib Theme

my_theme <- bslib::bs_theme(
  version = 5,                      # Bootstrap 5
  bootswatch = "flatly",            # Optional preset theme
  bg = "#ffffff",                   # Background color
  fg = "#2c3e50",                   # Foreground (text) color
  primary = "#2c3e50",              # Primary brand color
  secondary = "#95a5a6",            # Secondary color
  success = "#18bc9c",
  info = "#3498db",
  warning = "#f39c12",
  danger = "#e74c3c",
  base_font = bslib::font_google("Source Sans Pro"),
  heading_font = bslib::font_google("Source Sans Pro", wght = 600),
  code_font = bslib::font_google("Fira Code"),
  "navbar-bg" = "#2c3e50"
)

ui <- page_sidebar(
  theme = my_theme,
  title = "Themed App",
  # ...
)

Interactive theme editor during dev:

bslib::bs_theme_preview(my_theme)

→ App renders w/ consistent brand colors + fonts + BS5 components.

If err: fonts don't load → check net (Google Fonts needs it) or sys fonts: font_collection("system-ui", "-apple-system", "Segoe UI"). Theme vars don't apply → check theme passed to page fn.

Step 3: Cards + Columns Layout

ui <- page_sidebar(
  theme = my_theme,
  title = "Analysis Dashboard",
  sidebar = sidebar(
    width = 300,
    title = "Filters",
    selectInput("dataset", "Dataset", choices = c("iris", "mtcars")),
    sliderInput("sample", "Sample %", 10, 100, 100, step = 10),
    hr(),
    actionButton("refresh", "Refresh", class = "btn-primary w-100")
  ),

  # KPI row — non-filling
  layout_columns(
    fill = FALSE,
    col_widths = c(4, 4, 4),
    value_box(
      title = "Observations",
      value = textOutput("n_obs"),
      showcase = bsicons::bs_icon("table"),
      theme = "primary"
    ),
    value_box(
      title = "Variables",
      value = textOutput("n_vars"),
      showcase = bsicons::bs_icon("columns-gap"),
      theme = "info"
    ),
    value_box(
      title = "Missing",
      value = textOutput("n_missing"),
      showcase = bsicons::bs_icon("exclamation-triangle"),
      theme = "warning"
    )
  ),

  # Main content row
  layout_columns(
    col_widths = c(8, 4),
    card(
      card_header("Distribution"),
      full_screen = TRUE,
      plotOutput("main_plot")
    ),
    card(
      card_header("Summary"),
      tableOutput("summary_table")
    )
  )
)

Key primitives:

  • layout_columns() — responsive grid w/ col_widths
  • card() — container + optional header/footer
  • value_box() — KPI + icon + theme
  • layout_sidebar() — nested sidebar in cards
  • navset_card_tab() — tabbed cards

→ Responsive grid adapts to screen size.

If err: cols stack on wide screens → check col_widths sum = 12 (BS grid). Cards overlap → fill = FALSE on non-filling rows.

Step 4: Dynamic UI

server <- function(input, output, session) {
  output$dynamic_filters <- renderUI({
    data <- current_data()
    tagList(
      selectInput("col", "Column", choices = names(data)),
      if (is.numeric(data[[input$col]])) {
        sliderInput("range", "Range",
          min = min(data[[input$col]], na.rm = TRUE),
          max = max(data[[input$col]], na.rm = TRUE),
          value = range(data[[input$col]], na.rm = TRUE)
        )
      } else {
        selectInput("values", "Values",
          choices = unique(data[[input$col]]),
          multiple = TRUE
        )
      }
    )
  })

  # Conditional panels (no server round-trip)
  # In UI:
  # conditionalPanel(
  #   condition = "input.show_advanced == true",
  #   numericInput("alpha", "Alpha", 0.05)
  # )
}

→ UI updates based on user + data.

If err: dynamic UI flickers → conditionalPanel() (CSS) over renderUI() when poss. Dynamic inputs lose vals on re-render → session$sendInputMessage() restore state.

Step 5: Custom CSS/SCSS (Optional)

Beyond bslib theme vars:

# Inline CSS
ui <- page_sidebar(
  theme = my_theme,
  tags$head(tags$style(HTML("
    .sidebar { border-right: 2px solid var(--bs-primary); }
    .card-header { font-weight: 600; }
    .value-box .value { font-size: 2.5rem; }
  "))),
  # ...
)

# External CSS file (place in www/ directory)
ui <- page_sidebar(
  theme = my_theme,
  tags$head(tags$link(rel = "stylesheet", href = "custom.css")),
  # ...
)

SCSS + bslib:

my_theme <- bslib::bs_theme(version = 5) |>
  bslib::bs_add_rules(sass::sass_file("www/custom.scss"))

→ Custom styles applied w/o breaking bslib theming.

If err: custom CSS conflicts → use BS CSS vars (var(--bs-primary)) over hardcoded colors. Ensures theme changes propagate.

Step 6: A11y

# Add ARIA labels to inputs
selectInput("category", "Category",
  choices = c("A", "B", "C")
) |> tagAppendAttributes(`aria-describedby` = "category-help")

# Add alt text to plots
output$plot <- renderPlot({
  plot(data(), main = "Distribution of Values")
}, alt = "Histogram showing the distribution of selected values")

# Ensure sufficient color contrast in theme
my_theme <- bslib::bs_theme(
  version = 5,
  bg = "#ffffff",      # White background
  fg = "#212529"       # Dark text — 15.4:1 contrast ratio
)

# Use semantic HTML
tags$main(
  role = "main",
  tags$h1("Dashboard"),
  tags$section(
    `aria-label` = "Key Performance Indicators",
    layout_columns(
      # value boxes...
    )
  )
)

→ App meets WCAG 2.1 AA: contrast + keyboard + screen reader.

If err: test browser a11y audit (Lighthouse). Contrast → WebAIM checker. All interactive → keyboard-focusable.

Check

  • Layout OK on desktop + mobile
  • bslib theme consistent across components
  • Value boxes → correct themes + icons
  • Cards resize in grid
  • Custom CSS → BS vars, not hardcoded
  • Plots have alt text
  • Contrast ≥ WCAG AA (4.5:1 text)
  • Interactive → keyboard accessible

Traps

  • Mix old + new Shiny UI: No fluidPage() w/ bslib. Use page_sidebar() / page_navbar() / page_fillable() only.
  • Hardcoded colors: Use var(--bs-primary) over #2c3e50. Hardcoded → breaks on theme change.
  • Missing fill = FALSE: Value box + summary rows shouldn't stretch. fill = FALSE.
  • Google Fonts offline: Air-gapped → sys fonts or self-hosted font files over font_google().
  • Ignore mobile: Test browser responsive. layout_columns auto-stacks, custom CSS may not.

  • scaffold-shiny-app — initial setup + theme config
  • build-shiny-module — modular UI components
  • optimize-shiny-performance — perf-conscious render
  • review-web-design — visual review: layout + typo + color
  • review-ux-ui — usability + a11y review

GitHub Repository

pjt222/agent-almanac
Path: i18n/caveman-ultra/skills/design-shiny-ui
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

Related Skills

content-collections

Meta

This skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.

View skill

polymarket

Meta

This skill enables developers to build applications with the Polymarket prediction markets platform, including API integration for trading and market data. It also provides real-time data streaming via WebSocket to monitor live trades and market activity. Use it for implementing trading strategies or creating tools that process live market updates.

View skill

creating-opencode-plugins

Meta

This skill helps developers create OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It provides the plugin structure, event API specifications, and implementation patterns for JavaScript/TypeScript modules. Use it when you need to intercept, monitor, or extend the OpenCode AI assistant's lifecycle with custom event-driven logic.

View skill

sglang

Meta

SGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.

View skill