when a lot seems broken at once: check convey first

the most common support pattern we see: several parts of solstone appear to fail at the same time — the timeline, your entities, search, chat — and it looks like everything fell over.

almost always, it's one thing. convey is the service behind solstone's web ui and those app surfaces. when convey's runner stops, every app it serves goes dark together — so one failure looks like many.

check the health

from a terminal in your solstone:

journal doctor

this reports what's running and what isn't. if convey is down or restarting, that's your answer — the "broken agents" are just downstream of it.

bring convey back

journal restart-convey

then reload the web ui (by default http://localhost:5015) and the surfaces should return. run journal doctor again to confirm it's back up.

a note on sol vs. journal

day-to-day journal access is sol; host and service operations — like restarting convey — are journal. if sol restart-convey tells you the command isn't available there, that's expected: run it as journal restart-convey.

"401 unauthorized" is a different thing

if convey is up but you're seeing 401 unauthorized on convey calls, that's an authentication problem, not a crash — usually a missing or invalid device key, or a system clock that's drifted (auth is time-sensitive, so a wrong clock fails it). check which devices are registered to your journal:

journal observer list
journal observer status

and confirm your machine's clock is correct (date). if convey is also restarting, sort that out first.

still stuck?

if convey won't stay up, or doctor reports something else, file a request — sol call support create ... collects the diagnostics we need, or fill in the form at https://support.solstone.app/, or email support@solstone.app. include the output of journal doctor if you can.