Check your setup

When a machine will not come online, or a session refuses to start, or the app shows a computer you are sure is running, there is one command that tells you why:

happier status

It reads everything and changes nothing. You can run it whenever you want, on any machine, without wondering whether it will restart something mid-session.

What it looks at

happier status checks the whole chain between you and a working session:

• Your account — whether this machine is signed in, and whether the sign-in has expired.
• Relays — which relay profiles are configured, which is active, and whether the machine is registered against it. A machine can be signed in and still not registered, which is the usual reason a computer never appears in the app.
• Automatic startup — whether the background service is installed, running, pointing at the right channel, and not duplicated.
• The CLI itself — which channel it is on (happier, hprev, hdev), whether an update is waiting, and whether a running daemon was built from a different version than the CLI you just invoked.

It reports 27 distinct problems in those areas. Most are things you would never guess at: a background service left over from a channel you switched away from, two startup entries fighting over the same relay, a daemon still running from an older CLI, a local relay whose version has drifted from the client's.

Fixing what it finds

happier status never changes anything — it refuses --yes outright and tells you to use the repair command instead. When you are ready:

happier doctor repair

This walks the same findings and offers to fix each one, asking before it acts. Add --yes to accept every repair it can make without prompting, which is what you want in a script or over SSH:

happier doctor repair --yes

Add --report-only to see the plan without applying it. That is exactly what happier status runs underneath, so the two are interchangeable when you just want to look.

Some findings cannot be repaired automatically, and the report says so rather than pretending. Expired authentication is the common one: Happier can tell you the sign-in has lapsed, but only you can complete a login.

The background service

Most of what status reports is about automatic startup, because that is what keeps a machine reachable when no terminal is open. You can drive it directly:

happier service status      # the default if you pass nothing
happier service install
happier service start
happier service stop
happier service restart
happier service uninstall
happier service logs
happier service list
happier service paths
happier service repair

happier service logs is worth knowing before you need it — when a machine keeps dropping offline, the service log usually says why, and it is easier to read than reproducing the failure by hand.

Machine-readable output

Both commands can emit JSON, which is what you want in CI or a health check:

happier doctor --json
happier doctor repair --report-only --json

Clearing stuck processes

Rarely, a crashed run leaves processes behind that hold a port and stop the daemon from starting:

happier doctor clean

It reports how many it stopped. Reach for it when status reports a port in use or a conflicting manual daemon and a plain repair has not settled it.

Related

• CLI — the full command surface.
• Daemon — what the background process actually does.
• Onboarding — first-time setup.