A command is an NCL procedure that performs specific processing. Automation Services comes with a predefined set of commands.
Commands can be generic (that is, perform processing that is not related to a specific resource or service) or can apply to a class, to a type within a class, or to a specific resource or service. A more specific command has precedence over a less specific command.
You can define your own commands to perform site-specific processing by writing an NCL procedure and associating it with a command definition. Supplied command definitions can be modified to suit your requirements.
To execute commands programmatically from a batch process, exit procedure, console, or external application, use the $RMCALL API procedure.