How it works

From commit to production
in one command.

No YAML archaeology, no cluster to babysit. Point ReleaseMachine at your repo and the rele CLI handles the rest.

1

Initialize

Run rele init in your project. It detects your runtime from package.json, requirements.txt, go.mod, or Gemfile, writes a tiny rele.yml, and synthesizes a Dockerfile if you do not have one.

$ rele init # detected: node 20 (package.json) # wrote rele.yml ready to deploy.
2

Deploy

One command builds your image, pushes it to a private registry, rolls it out on the cluster, and streams the logs back to your terminal. Most deploys finish in under ten seconds.

~/my-app - rele

        
3

Ship, and roll back if you need to

Every deploy is tagged with its git SHA and kept for 90 days. If a release goes sideways, roll back to any point in history with a single command. No dashboards to dig through.

$ rele deploy --ref a1b2c3d # rolling back to a1b2c3d ✓ live at my-app.rel.run 1.6s

What you get, every deploy

Zero config

Convention over configuration. We detect your stack and build it. Bring a Dockerfile if you want full control.

Speed you feel

Pre-warmed builders and layer caching keep the build-push-deploy loop tight, so shipping feels instant.

Logs and status

Tail live output with rele logs and check what is running with rele status.

$ rele logs --follow

Secrets, handled

Set env vars with rele env set. They are encrypted at rest and injected at deploy time, never baked into images.

$ rele env set DATABASE_URL=...

Auto-deploy on push

Run rele ci init to wire up GitHub Actions. Every push to main ships automatically.

Bring your own database

Connect a managed Postgres or MySQL you already trust. No lock-in, no proprietary data store.

Want the full command list? Preview the CLI reference.

Ship your next app in seconds.

ReleaseMachine is in private development. Join the waitlist and we’ll tell you the moment it’s live.