Run your current directory in Linux, from macOS or Windows.
cd ~/code/my-project
avr # interactive Linux shell, same project
avr npm test # run one command in Linux
The mental model is the current directory plus the operating environment you pick. There is no machine to name, no mounts to configure and no SSH to set up. avar runs Lima virtual machines on macOS and WSL 2 distributions on Windows, and the commands are the same on both.
Installing avar and getting to a first Linux shell take one page, and that page is the README:
This site holds what does not fit on that page.
| Page | What it covers |
|---|---|
| Commands | A page per command: its help, examples, exit statuses and errors |
| Syntax | The command line, .avr.toml, config.toml, and how they combine |
| Environments | The distributions, releases and architectures avar runs, on each host |
| Platform notes | What Linux can see, the Windows filesystem boundary, --native-fs and avr sync |
| Design decisions | Why avar works the way it does, and what each choice costs |
| Troubleshooting | Common errors, what they mean, and what to do |
avar is developed spec-first, and every change arrives as a pull request. CONTRIBUTING.md explains how to propose one, and docs/lessons.md records the mistakes that changed how the project is worked on.