A Databricks App powered by AppKit, featuring React, TypeScript, and Tailwind CSS.

Enabled plugins: {{- if .plugins.analytics}}

• Analytics -- SQL query execution against Databricks SQL Warehouses {{- end}} {{- if .plugins.lakebase}}
• Lakebase -- Fully managed Postgres database for transactional (OLTP) workloads on Databricks {{- end}} {{- if .plugins.genie}}
• Genie -- AI/BI Genie conversational interface for natural language data queries {{- end}}
• Server -- Express HTTP server with static file serving and Vite dev mode

• Node.js v22+ and npm
• Databricks CLI (for deployment)
• Access to a Databricks workspace

For local development, configure your environment variables by creating a .env file:

cp .env.example .env

Edit .env and set the environment variables you need:

DATABRICKS_HOST=https://your-workspace.cloud.databricks.com
DATABRICKS_APP_PORT=8000
# ... other environment variables, depending on the plugins you use

{{- if .plugins.lakebase}}

The Lakebase plugin requires additional environment variables for PostgreSQL connectivity. To learn how to configure the Lakebase plugin, see the Lakebase plugin documentation. {{- end}}

The Databricks CLI requires authentication to deploy and manage apps. Configure authentication using one of these methods:

Interactive browser-based authentication with short-lived tokens:

databricks auth login --host https://your-workspace.cloud.databricks.com

This will open your browser to complete authentication. The CLI saves credentials to ~/.databrickscfg.

Use multiple profiles for different workspaces:

[DEFAULT]
host = https://dev-workspace.cloud.databricks.com

[production]
host = https://prod-workspace.cloud.databricks.com
client_id = prod-client-id
client_secret = prod-client-secret

Deploy using a specific profile:

databricks bundle deploy --profile production

Note: Personal Access Tokens (PATs) are legacy authentication. OAuth is strongly recommended for better security.

npm install

Run the app in development mode with hot reload:

npm run dev

The app will be available at the URL shown in the console output.

Build both client and server for production:

npm run build

This creates:

• dist/server.js - Compiled server bundle
• client/dist/ - Bundled client assets
• .appkit/ - TypeScript type definitions cache (committed to git; do not ignore)

Run the production build:

npm start

There are a few commands to help you with code quality:

# Type checking
npm run typecheck

# Linting
npm run lint
npm run lint:fix

# Formatting
npm run format
npm run format:fix

Update databricks.yml with your workspace settings:

targets:
  default:
    workspace:
      host: https://your-workspace.cloud.databricks.com

Make sure to replace all placeholder values in databricks.yml with your actual resource IDs.

Deploy and start the app with a single command:

databricks apps deploy

databricks apps deploy validates the project, deploys it, starts the app, and prints its URL.

1. Configure the production target in databricks.yml
2. Deploy to production:

databricks apps deploy -t prod

Restarting a stopped app: apps stop after a period of inactivity. To start one again without redeploying, run databricks apps start <APP_NAME>.

* client/          # React frontend
  * src/           # Source code
  * public/        # Static assets
* server/          # Express backend
  * server.ts      # Server entry point
  * routes/        # Routes
* shared/          # Shared types
{{- if .plugins.analytics}}
* config/          # Configuration
  * queries/       # SQL query files
{{- end}}
* databricks.yml   # Bundle configuration
* app.yaml         # App configuration
* .env.example     # Environment variables example

• Backend: Node.js, Express
• Frontend: React.js, TypeScript, Vite, Tailwind CSS, React Router
• UI Components: Radix UI, shadcn/ui
• Databricks: AppKit SDK