Yes please. I understand the arguments against, but respectfully disagree - it's a pita to have to reverse engineer the key setup and copy it into a ton of different repo workflows. And then there is the possibility of future changes. I'd like the setup-go action to "just handle all of that". i.e. I think the complexity should be shifted away from users.

Can this PR get reviewed? A use case I have for this is to be able to speed up deployments. Currently waiting for the go cache cleanup to push at the end adds a noticeable delay and the deployment build is not going to produce a useful cache that doesn't already exist. I am looking to see if it is spending time caching unnecessary things and if there are settings for that, but would be nice to avoid altogether in this case.

This would be great to get merged. @HarithaVattikuti ?

This would be a great (and simple) addition.