Configuration, flags, and the safety model.
Everything you need to run rootuno on a host: how it's configured, how licensing and the pooled key work, and exactly what triggers a confirmation prompt.
Getting started
rootuno runs as root on the host it manages — one process, one binary, no agent fleet to deploy. If your host doesn't have it installed yet, reach out at [email protected] or check the onboarding email from your dashboard signup.
Activate with a license
rootuno runs on a license key — there's no key of your own to manage. Run rootuno -license <key> once: rootuno validates the key against the dashboard, saves it locally, and from then on activates automatically to get rootuno's pooled key.
# first run, activating with a license key $ rootuno -license ru_live_xxxxxxxxxxxxxxxx ✓ license valid — plan: free ✓ activated — pooled key cached locally # every run after that just works $ rootuno > what's using all the memory on this box?
Configuration
The only thing you configure is your license key. Set it with rootuno -license <key> (saves it to ~/.config/rootuno/config.json), or via the ROOTUNO_LICENSE_KEY environment variable. Everything else — model, API endpoint, request behavior — is decided by rootuno.
CLI flags
| Flag | Effect |
|---|---|
-yes, -y |
Auto-approve risky actions without prompting. Use with care — this removes the human-in-the-loop check entirely for that run. |
-license <key> |
Validate the key against the dashboard and save it as license_key in config.json. From then on the CLI activates against the dashboard for the pooled key automatically. |
-message <text>, -m |
Run a single instruction non-interactively and exit, instead of dropping into the REPL. Same effect as passing the instruction as plain arguments — useful when you want an explicit, quoted flag, e.g. from a cron job. |
Licensing & the pooled key
The dashboard at dashboard.rootuno.com is where license keys, plans, and usage live. A valid license activation hands the CLI rootuno's own pooled key, encrypted and cached locally so brief dashboard outages don't stop you working. There's no bring-your-own-key option — a license is required to run rootuno.
New accounts start on the free plan.
Safety model
rootuno draws a hard line between "look around" and "change something irreversible," and it draws that line automatically — not by asking the model to decide when to be careful.
Destructive-command detection
Shell commands are pattern-matched for destructive intent before they run: recursive deletions, disk operations, reboots and shutdowns, firewall flushes, and service or package removal all match.
Protected write paths
File writes to sensitive locations require confirmation regardless of how the change was requested, including /etc, /boot, systemd unit files, and /root/.ssh.
The confirmation prompt
A matched action stops and waits for an explicit y/N in the terminal. Decline, and rootuno explains what it wanted to do and why — it doesn't retry silently or route around the refusal. Pass -yes at launch to skip prompts for the whole run, at your own risk.
Example session
A typical run: rootuno investigates first, proposes the fix, and only asks permission for the part that actually matters.
> the disk is almost full on this box, sort it out running: df -h / running: du -sh /var/log/* | sort -rh | head running: journalctl --disk-usage ⚠ agent wants to run: journalctl --vacuum-size=200M proceed? [y/N] y running: journalctl --vacuum-size=200M ✓ freed 4.1G — journal logs were the culprit (was 91% full, now 58%). rotated old logs, left everything else untouched.
Ready to point it at a real problem?
Activate a license from the dashboard and run rootuno directly.