Divekit CLI for Beginners
Learn how to install Divekit CLI, configure authentication, create your first origin repository, add a simple individualization, and run a first distribution safely.
1 hr
13 Lections
Updated July 2026
Torben Buck
Software Developer
This course is a practical introduction to Divekit CLI.
You will work through one consistent example: a small programming assignment called vehicle-workshop.
By the end of the course, you will be able to set up Divekit on your machine, connect it to a GitLab host, prepare a first origin repository, add a simple individualization, and distribute individualized repositories.
Target audience
This course is for people who:
- have never used Divekit before
- can work with a terminal and edit files
- want to learn the full beginner workflow from local setup to first distribution
Prerequisites
Before you begin, make sure you have:
- a Linux, macOS, or Windows machine
- access to a GitLab instance
- permission to create groups and repositories on the target GitLab instance
- enough confidence to read and edit small JSON files during the course
You will need a Personal Access Token with api scope later during the authentication step.
The course shows that part when it becomes relevant.
Learning outcomes
After completing this course, you will be able to:
- explain the role of an origin repository and a distribution
- install Divekit CLI and verify your setup
- configure a GitLab host and token with Divekit
- create a new Divekit origin repository with
divekit init - review the generated
config.jsonfor a first rollout - create a minimal
individualization.json - provide members and understand how they map to repositories
- perform a first distribution and inspect the generated state
- clean up test rollouts with
divekit purge
Recommended way to work through the course
Read each lesson in order and do the exercises directly in a real terminal.
When a lesson asks you to inspect a generated file, actually open that file and read it.
Beginner mistakes often come from not knowing what a normal file or a normal command result should look like.