Skip to main content

Getting started

This guide gets you to a running webapp/client on http://localhost:3000, talking to the orchestration-cluster backend on http://localhost:8080.

Prerequisites​

  • Node.js via fnm (recommended) or nvm. The repo pins Node in webapp/client/.nvmrc — fnm use / nvm use reads it for you.
  • Java — the backend builds and runs with Maven. Install via SDKMAN or Homebrew (brew install openjdk@21). Check the current version on the project.
  • Docker — used by make env-up to run Elasticsearch, and by Playwright's containerized browser for visual-regression tests.
  • make — wraps backend startup (make env-up / make env-down) and VS Code config sync (make vscode-sync-all). Pre-installed on macOS and most Linux; Windows users install via WSL or choco install make.

Clone & install​

git clone https://github.com/camunda/camunda.git
cd camunda/webapp/client
fnm use            # or: nvm use
npm ci

The workspace install covers all packages/* and apps/*.

Run the backend​

From webapp:

make env-up

This boots Elasticsearch (Docker) and the orchestration-cluster backend on http://localhost:8080, seeded with a demo / demo user. Tear it down with:

make env-down

Run the frontend​

From webapp/client:

npm run dev:oc

The dev server opens http://localhost:3000. Vite proxies /v2, /login, and /logout to :8080, so the frontend talks to the backend you started above.

Verify​

  • The dev page renders in the browser.
  • npm run lint exits clean.
  • npm run typecheck exits clean.

Optional: VS Code setup​

From the repo root:

make vscode-sync-all

Merges the repository's MCP and editor settings into your VS Code config.