Skip to content

Getting started

Seven pages, best read in order. All you need is a terminal and access to a model provider, and for the last page, a hub that someone runs for your team.

  • Installation: what you need, go install, building from source, the container image and shell completion. At the end you have a working cloop binary.
  • Your first project: a complete first run. You set a goal, read the plan cloop makes from it, run it, watch it, change it and stop it. If you only read one page, read this one.
  • How cloop works: what a task is, how cloop picks the next one and knows when it is finished, what happens when a task fails, and what auto-evolve does once the plan is done.
  • Choosing and configuring a provider: the model backends cloop can use, what each is good for, how cloop decides which provider and model to use, and the environment variables that override that.
  • The web dashboard: running cloop ui, what each screen shows, and what to set up before you let anyone else reach it.
  • A walkthrough of the dashboard: a first project again, this time in the browser, with screenshots. You choose where work may run, create a project that runs there, add tasks and start it.
  • Working from a terminal: signing your terminal in to a hub with cloop login. The terminal prints a code, you approve it in the browser you are signed in with, and the terminal gets its own session. Then cloop remote for projects, tasks, runs and live logs, cloop task attach, and signing out.

After that

The pages above assume one machine and one person. The rest of the documentation covers what changes when there are more:

↩ documentation map