Autar CLI

Get started with the Autar CLI

Download the CLI, authenticate, connect your source provider, and access your repositories from the terminal.

1
Install

Download and install the current release.

macOS and Linux
curl -fsSL --proto '=https' https://api.autar.ai/cli/download/install.sh | sh
Windows (PowerShell)
irm -MaximumRedirection 0 https://api.autar.ai/cli/download/install.ps1 | iex

On macOS and Linux, add ~/.local/bin to your PATH if it is not already there. The Windows installers add their directory for you.

Update the CLI.

autar update
2
Sign in Interactive use
autar login

Sign in with your browser.

Automation

Create a personal access token in Autar API token settings and set it in the environment.

export AUTAR_TOKEN="your-token"
3
Connect a provider

Connect GitHub and list your repositories.

autar providers connect github
autar providers sync github
autar repos

Replace github with gitlab or bitbucket when appropriate.

Release notes

Autar CLI v0.1.5 Release Notes

This production release brings a new Automations workspace, commit-history intelligence, and better troubleshooting to the terminal.

What's new

  • New Automations TUI: Discover maintenance opportunities, run tasks across selected repositories, review proposed changes, and track execution history.
  • Commit history intelligence: Find recurring maintenance patterns, inspect supporting commits, estimate future repetition, and identify automation candidates. Export reports as text, Markdown, or JSON.
  • Local ZIP scans: Upload and analyze a local repository archive with autar scan --zip.
  • Estate and Fleet scoping: Save your working scope or select it per command.
  • Verified self-update: Check for and install releases with autar update, including release-channel and checksum verification.

Fixes

  • Improved browser login resilience, including recovery from transient server errors.
  • Improved cancellation, pagination, filtering, task-state reconciliation, and review handling.
  • Made ZIP upload confirmation and cleanup more reliable.
Scan report formats

text: compact terminal summary, the default.

md: full Markdown report.

json: analysis data for automation.

autar report <job-id>
autar report <job-id> --output md
autar report <job-id> --output json

Running this across more than one repo?

See Enterprise Fleet