Introduction

Kuppit deploys applications from GitHub without asking you to assemble cloud infrastructure. You connect a repository, Kuppit works out how to build it, and every push becomes a deployment with a public hostname, logs, variables and the resources it depends on.

These docs answer three questions:

• How do I do something? Deploy an application, add a database, set a variable, connect two resources. Start with the Quickstart.
• How does Kuppit work? Projects, environments, resources, deployments and the graph between them. Start with The Kuppit model.
• Something broke. What now? When a deployment fails says what each failure means and what to do; Logs is where to read what happened. A dedicated troubleshooting library follows.

What you need

• A GitHub account. Kuppit signs you in with it and reads repositories through a GitHub App you install.
• A repository containing either a Dockerfile or a Node.js application Kuppit can build automatically: Nuxt, Next.js, Express, or anything with a start script. See Web services.
• A Kuppit account. Kuppit is in private alpha. Sign in at app.kuppit.run with GitHub and your account is held for approval.

The shape of every deployment

1. Install the GitHub App on a repository.
2. Create a project from it. Kuppit analyses the repository and proposes a web service.
3. Deploy. Kuppit builds a container image, starts it, checks that it is healthy, then moves traffic to it.
4. Add resources such as PostgreSQL and connect them to your services through variables.
5. Push again. Every push to the deployed branch creates a new deployment; every pull request gets its own preview environment.

The Quickstart walks through all five with a real repository.