Using the dfeshiny Deploy Template

When developing DFE Shiny dashboards, it’s common to want a consistent, simple, and secure way to deploy applications to platforms such as ShinyApps.io or the new DfE hosting platform. This article shows how to use the dfeshiny reusable workflow template for GitHub Actions, with parameters defined in a separate YAML file to minimize duplication and speed up development.

Overview

Have a deploy_parameters.yaml: Store all your deployment parameters, including the app name and deployment target.
Minimal Local Workflow: Reference the remote reusable workflow (stored in another repository) from your local .github/workflows/deploy.yaml.
Secrets Management: Ensure your repository has the required secrets (e.g., SHINYAPPS_TOKEN and SHINYAPPS_SECRET) set up under Settings > Secrets and variables.

1. Create a `deploy_parameters.yaml`

At the root of your project, add a file called deploy_parameters.yaml.
Below is an example with optional parameters:

dashboard_name: "my-cool-dashboard"
deploy_target: "shinyapps"

dashboard_name: Your application’s name on ShinyApps (or other hosting platform).
deploy_target: Where you want to deploy (e.g., "shinyapps").

2. Set Up Your Local Workflow

In your local .github/workflows/ folder, create or edit a file—commonly named deploy.yaml. This file defines only the trigger conditions (which branches to watch) and then calls the remote reusable workflow.

Here is a minimal example:

name: Deploy Dashboard

on:
  push:
    branches:
      - main
      - development

jobs:
  deploy:
    uses: dfe-analytical-services/dfeshiny/.github/workflows/deploy-dashboard-template.yaml@main
    with:
      parameter_file: deploy_parameters.yaml

on: push… ensures this workflow runs whenever code is pushed to main or development.
uses: dfe-analytical-services/dfeshiny-deploy-templates… points to the remote repo and file containing the actual deployment steps.
with: parameter_file: deploy_parameters.yaml tells the remote workflow which file to parse for parameters.

3. Secrets and Permissions

Before your deployment can succeed, you need to set the relevant secrets in your local repository. These secrets are not shared automatically across repos, so you must create them in your project’s Settings > Secrets and variables > Actions:

SHINYAPPS_TOKEN
SHINYAPPS_SECRET

4. How the Template Works

Once you commit your deploy.yaml and deploy_parameters.yaml, the GitHub Actions runner will:

Check Out the Code – Pulls your repository code and the deploy_parameters.yaml.
Parse Parameters – Uses a tool like yq to read deploy_parameters.yaml:
- dashboard_name
- deploy_target
Restore R Dependencies – If your project uses renv, it runs renv::restore() to ensure your local library is consistent.
Deploy – Calls the appropriate deployment method based on deploy_target.
- For ShinyApps.io, it uses rsconnect::deployApp().

Summary

Using a reusable GitHub Actions workflow template:

Reduces complexity of dashboards deployments - you only maintain a minimal local config .
Ensures consistent deployment practices across your dashboards.
Makes it easy to update deployments across your dashboards with minimal changes in each repository.

If you have questions about credentials or advanced usage, contact the Explore Education Statistics Platforms team at explore.statistics@education.gov.uk.

Whilst following the instructions in this article will help you set up the deployment of your dashboard, you will still need to seek approval to publish from the Explore Education Statistics team and ask for them to set up the secrets management on GitHub for your deployment to go through.