Uninstall

Use reploy uninstall to remove installed service wiring and stop the Docker objects Reploy created for an installed deployment.

When the deployment directory still exists, uninstall from the directory:

sudo reploy uninstall --from /opt/example

User-scope Docker-managed permanent installs are uninstalled from the installed target without sudo:

reploy uninstall --from "$PWD/example-installed"

On Linux, if the directory was already deleted, uninstall by service name:

sudo reploy uninstall --service-name example

List known Reploy services:

reploy services list

Remove the target directory as part of uninstall:

sudo reploy uninstall --from /opt/example --remove-dir

The service-name flow is intended for recovery when a target directory was manually deleted but Docker or system service state still exists. Docker Desktop-backed macOS uninstall requires the installed deployment state at --from. Recovery trusts only an exact, regular systemd unit marked as managed by Reploy, verifies its recorded service, target, and Compose project, and refuses to run if the target directory still exists. Because the deleted directory also deletes its run queue, admission mode flags have no additional effect on this recovery-only path.

By default, uninstall fails while app commands or shell sessions are active or waiting. Use --wait to join the queue, --drain to let active runs finish and cancel queued runs, or --force to stop active runs and cancel queued runs.