Recreate `secret` command in yargs

[GeorgeCadwallader]

Problem to solve

We need to recreate the secret command in version 2 of the app.

Proposal

This command should work the same as it did in version 1.

We're not looking at supporting anymore OS types with this command, purely to replicate the same functionality from version 1 to 2.

The idea really is that we will advise people pass the token in manually by using the CT_TOKEN variable. As they would do when integrating Conventional Tools in CI.

[GeorgeCadwallader]

@AdeAttwood What do we think about this one?

Been playing around with this and it might be best to just leave this one I think.

As we discussed previously, we want to push the use of passing CT_TOKEN in instead. Also with pushing the idea that Conventional Tools should be used in CI (when CT_TOKEN is required).

Also with not wanting to use execa (which is how this is done in 1.x) and child processes I think it might be best to leave this.

[AdeAttwood]

I think it might be best to leave this.

I don't mind. If we are on the fence about removing this feature in V2 then we should leave if off the list, if it becomes a probity we can add it back in. I don't think this is a big feature and will not be a significant amount of work to build in. We will still need to abstract the credentials logic, maybe bundle it into source control host interface like we have the source control.

[GeorgeCadwallader]

I think it might be best to leave this.

I don't mind. If we are on the fence about removing this feature in V2 then we should leave if off the list, if it becomes a probity we can add it back in. I don't think this is a big feature and will not be a significant amount of work to build in. We will still need to abstract the credentials logic, maybe bundle it into source control host interface like we have the source control.

Cool, I'll move this into the Backlog milestone and we can move back in to 2.x if we think it's a requirement.