Not quite "global", but in the current beta (and of course the future release), at the profile level you can use the "Execute a command each time a command phrase is recognized" option to have a command execute whenever a spoken command is recognized.
Said command can then use an action (E.G. "Play a Sound") for audible feedback.