Skip to main content
The runner is a signed binary that executes your agents’ work — cloning, editing, running tests, opening pull requests — on a Mac you control. See Where things run for what that split buys you; this page is the mechanics of getting one running.

Before you start

An Apple Silicon Mac

Running macOS 13 or later. Intel is not supported.

Someone able to stay logged in

The runner needs a login session — see The login session below before you pick which Mac.

An OpenRouter account

Model spend for the actual work is billed to you directly, on a key that never leaves this machine.

An enrollment token

From Runners in your workspace sidebar → Enroll a runner. Owner or admin only, single-use, and expires in fifteen minutes.

The login session

The runner installs as a LaunchAgent, not a LaunchDaemon — it runs only while someone is logged in. This is deliberate: the Keychain holding your OpenRouter key, and your coding tool’s own login, both live in a user’s session, not in a boot-time daemon with none. The recommended setup is a dedicated macOS user with auto-login, on a machine that is not rebooted unattended. If your policy requires FileVault, budget for someone unlocking the Mac after every reboot — nothing is lost while it is down, cases simply wait. See Known gaps for the honest version of this trade-off.

Sleep

A Mac that goes to sleep takes its agent offline with it: the agent can’t answer until someone wakes the Mac. The dashboard’s runner says Asleep when the Mac’s sleep reached us before it froze; more often the Mac can only say so once it wakes, and then the runner’s Timeline marks the time it was away as asleep. Keep this Mac awake while its agent is online is in the desktop app’s Settings → General. It works with the app closed, except for an agent box on Docker (Macs before macOS 26), where it holds only while the app is open. The display can still turn off. If you turn it off on a Mac set to sleep while plugged in, the app warns you. A closed lid always sleeps a MacBook (only an administrator’s sudo pmset changes that, and we never do); for an agent that must always be reachable, use a Mac mini or another desktop Mac.

Set it up

Confirming it worked

The runner appears on the Runners page as online within a few seconds. If it does not, and you used the command line:
The desktop app’s monitor shows the same doctor output without a terminal.

Next: give it something to work with

Enrolling gets the machine online; an agent still needs a coding tool signed in on it. From the agent’s page in the dashboard, its readiness panel checks this runner directly — which tools are installed, which are logged in, which repositories it can actually clone — and tells you the exact command to run for whatever is missing. The desktop app’s setup page checks the same things in place.

Getting a newer version

Desktop app

The desktop app updates itself, and the runner inside it comes along. A while after it opens, and every four hours after that, the app checks for a new release and downloads it in the background. When it is ready, the sidebar shows Update ready · Restart with the new version, and you get one notification for it. Click it to restart into the new version, or just quit; the update installs then either way. Restarting the app never interrupts work. The runner keeps running while the app restarts, and the new app moves it onto the new runner once no step is running. A runner in the isolated box moves onto the new version’s box image the same way: the box restarts once no step is running, keeping its sign-ins and settings. To check right away, choose Agentick → Check for Updates… from the menu bar. Settings → General shows the app’s version beside the runner’s. The app can only replace itself from your Applications folder. If you open it straight from the downloaded disk image, it says Move Agentick to Applications to get updates automatically instead; drag it to Applications and open it from there. If you installed with Homebrew, the app keeps itself current, so brew upgrade leaves it alone. brew upgrade --greedy --cask agentick-desktop still works if you prefer it. A runner that ships inside the app never replaces itself, even when your workspace offers it a newer release: its log says the update is held because the app updates it, and agentick-runner self-update tells you to update the app instead. An app from before automatic updates (version 0.1.16 or older) needs one last update by hand: brew upgrade --cask agentick-desktop, or download the latest .dmg from the releases page and drag it over the old copy in Applications.

Command line

Self-updating is designed into the runner binary: it checks your workspace for a newer release, drains in-flight work first, verifies the new binary’s signature twice, and swaps it in atomically, rolling back if the new one fails to start. It is not switched on yet. Until it is, get the newest version the same way you installed it:
Restart it afterward so it picks up the new binary:
We will say here, plainly, once fleet-wide automatic updates for the command line runner are live, rather than let you discover their absence from a stale version.

Uninstall

Dragging Agentick to the Trash does not stop the runner: the box (or the runner on the Mac) keeps running in the background. Remove it first, from the app:
  1. Settings → General → Remove from this Mac.
  2. Leave Also revoke it in your workspace ticked (needs an Admin) so its credential stops working; tick Also delete the case folders if you want ~/.agentick/runner/worktrees gone too.
  3. Remove. The box, its sign-ins and settings, its background job and its volume are deleted (or the runner on the Mac, its LaunchAgent and its Keychain items). Then quit Agentick and drag it to the Trash.
A revoked runner keeps its name in the workspace until you delete it on the Runners page. Installed from the command line, remove it by hand:
and revoke it from the Runners page — deleting the binary stops it running, but revoking is what stops its credential from working at all.