Control Stacks (up/down/update)

Start, Stop and Update Stacks

/compose-up — start a stack

/compose-up [project] starts a project with up -d. Without an argument the command shows the stopped projects; with several you must name one explicitly. Before starting, it checks whether a .env exists and is readable — an unreadable .env would silently turn variables into empty values and therefore aborts. After starting, the health status of the services is printed.

/compose-up app-stack

/compose-down — stop a stack

/compose-down <project> [--remove] stops a stack. By default only compose stop runs — the project stays in the index as exited and can be restarted instantly with /compose-up. With --remove a full compose down is performed (containers and network are removed).

/compose-down app-stack            # stop only
/compose-down app-stack --remove   # full teardown

If the project is on the critical_compose_projects whitelist in the profile, the command refuses to stop it — unless you set the environment variable SM_CONFIRM_CRITICAL=yes. This keeps important stacks (database, monitoring) from being halted by accident.

/compose-update — update images

/compose-update <project> runs pull and up -d as one atomic chain (if the pull fails, nothing is recreated). It then shows an image change summary (which service moved to which new image) and the health status.