scaffold-shiny-app

name	scaffold-shiny-app
description	Scaffold a new Shiny application using golem (production R package), rhino (enterprise), or vanilla (quick prototype) structure. Covers framework selection, project initialization, and first module creation. Use when starting a new interactive web application in R, creating a dashboard or data explorer prototype, setting up a production Shiny app as an R package with golem, or bootstrapping an enterprise Shiny project with rhino.
license	MIT
allowed-tools	Read Write Edit Bash Grep Glob
metadata	{"author":"Philipp Thoss","version":"1.0","domain":"shiny","complexity":"basic","language":"R","tags":"shiny, golem, rhino, scaffold, web-app, reactive"}

Scaffold Shiny App

Create a new Shiny application with production-ready structure using golem, rhino, or vanilla scaffolding.

When to Use

• Starting a new interactive web application in R
• Creating a dashboard or data explorer prototype
• Setting up a production Shiny app as an R package (golem)
• Bootstrapping an enterprise Shiny project (rhino)

Inputs

• Required: Application name
• Required: Framework choice (golem, rhino, or vanilla)
• Optional: Whether to include module scaffolding (default: yes)
• Optional: Whether to use renv for dependency management (default: yes)
• Optional: Deployment target (shinyapps.io, Posit Connect, Docker)

Procedure

Step 1: Choose Framework

Evaluate the project requirements to select the appropriate framework:

Framework	Best For	Structure
golem	Production apps shipped as R packages	R package with DESCRIPTION, tests, vignettes
rhino	Enterprise apps with JS/CSS build pipeline	box modules, Sass, JS bundling, rhino::init()
vanilla	Quick prototypes and learning	Single app.R or ui.R/server.R pair

Expected: Clear framework decision based on project scope and team needs.

On failure: If unsure, default to golem — it provides the most structure and can be simplified later. Vanilla is only appropriate for throwaway prototypes.

Step 2: Scaffold the Project

Golem Path

golem::create_golem("myapp", package_name = "myapp")

This creates:

myapp/
├── DESCRIPTION
├── NAMESPACE
├── R/
│   ├── app_config.R
│   ├── app_server.R
│   ├── app_ui.R
│   └── run_app.R
├── dev/
│   ├── 01_start.R
│   ├── 02_dev.R
│   ├── 03_deploy.R
│   └── run_dev.R
├── inst/
│   ├── app/www/
│   └── golem-config.yml
├── man/
├── tests/
│   ├── testthat.R
│   └── testthat/
└── vignettes/

Rhino Path

rhino::init("myapp")

This creates:

myapp/
├── app/
│   ├── js/
│   ├── logic/
│   ├── static/
│   ├── styles/
│   ├── view/
│   └── main.R
├── tests/
│   ├── cypress/
│   └── testthat/
├── .github/
├── app.R
├── dependencies.R
├── rhino.yml
└── renv.lock

Vanilla Path

Create app.R:

library(shiny)
library(bslib)

ui <- page_sidebar(
  title = "My App",
  sidebar = sidebar(
    sliderInput("n", "Sample size", 10, 1000, 100)
  ),
  card(
    card_header("Output"),
    plotOutput("plot")
  )
)

server <- function(input, output, session) {
  output$plot <- renderPlot({
    hist(rnorm(input$n), main = "Random Normal")
  })
}

shinyAppui server

Expected: Project directory created with all scaffolding files.

On failure: For golem, ensure the golem package is installed: install.packages("golem"). For rhino, install from GitHub: remotes::install_github("Appsilon/rhino"). For vanilla, ensure shiny and bslib are installed.

Step 3: Configure Dependencies

Golem/Vanilla

# Initialize renv
renv::init()

# Add core dependencies
usethis::use_package("shiny")
usethis::use_package("bslib")
usethis::use_package("DT")         # if using data tables
usethis::use_package("plotly")     # if using interactive plots

# Snapshot
renv::snapshot()

Rhino

Dependencies are managed in dependencies.R:

# dependencies.R
library(shiny)
library(bslib)
library(DT)

Expected: All dependencies recorded in DESCRIPTION (golem) or dependencies.R (rhino) and locked with renv.

On failure: If renv::init() fails, check write permissions. If packages fail to install, check R version compatibility.

Step 4: Create First Module

Golem

golem::add_module(name = "dashboard", with_test = TRUE)

This creates R/mod_dashboard.R and tests/testthat/test-mod_dashboard.R.

Rhino

Create app/view/dashboard.R:

box::use(
  shiny[moduleServer, NS, tagList, h3, plotOutput, renderPlot],
)

#' @export
ui <- function(id) {
  ns <- NS(id)
  tagList(
    h3("Dashboard"),
    plotOutput(ns("plot"))
  )
}

#' @export
server <- function(id) {
  moduleServer(id, function(input, output, session) {
    output$plot <- renderPlot({
      plot(1:10)
    
  

Vanilla

Add module functions to a separate file R/mod_dashboard.R:

dashboardUI <- function(id) {
  ns <- NS(id)
  tagList(
    h3("Dashboard"),
    plotOutput(ns("plot"))
  )
}

dashboardServer <- function(id) {
  moduleServer(id, function(input, output, session) {
    output$plot <- renderPlot({
      plot(1:10)
    })
  })
}

Expected: Module file created with UI and server functions using proper namespacing.

On failure: Ensure the module uses NS(id) for all input/output IDs in the UI function. Without namespacing, IDs will collide when the module is used multiple times.

Step 5: Run the Application

# Golem
golem::run_dev()

# Rhino
shiny::runApp()

# Vanilla
shiny::runApp("app.R")

Expected: Application launches in the browser without errors.

On failure: Check the R console for error messages. Common issues: missing packages (install them), port already in use (specify a different port with port = 3839), or syntax errors in UI/server code.

Validation

• Application directory has correct structure for chosen framework
• shiny::runApp() launches without errors
• At least one module is scaffolded with UI and server functions
• Dependencies are recorded (DESCRIPTION or dependencies.R)
• renv.lock captures all package versions
• Module uses NS(id) for proper namespace isolation

Common Pitfalls

• Choosing vanilla for production: Vanilla structure lacks testing infrastructure, documentation, and deployment tooling. Use golem or rhino for anything beyond prototypes.
• Missing namespace in modules: Every inputId and outputId in a module UI must be wrapped with ns(). Forgetting this causes silent ID collisions.
• golem without devtools workflow: golem apps are R packages. Use devtools::load_all(), devtools::test(), and devtools::document() — not source().
• rhino without box: rhino uses box for module imports. Don't fall back to library() calls — use box::use() for explicit imports.

Related Skills

• build-shiny-module — create reusable Shiny modules with proper namespace isolation
• test-shiny-app — set up shinytest2 and testServer() tests
• deploy-shiny-app — deploy to shinyapps.io, Posit Connect, or Docker
• design-shiny-ui — bslib theming and responsive layout design
• create-r-package — R package scaffolding (golem apps are R packages)
• manage-renv-dependencies — detailed renv dependency management