Support primitive Dry-Run workflow

[WitoDelnat]

I was playing around a bit with the extension and I'd like to get it working with a tool like Kustomize or Helm. These tools often run in the CLI so part one of the challenge is getting the content within a VSCode model. I discovered that you can use helm template blue-helm-cert-manager | code - to open it in a tmp file (e.g. the output for that command is Reading from stdin via: /var/folders/7_/54m525216j753qdynl70gjjh0000gn/T/code-stdin-MkG for me). However, since this is not part of my workspace it will not get validated.

On a deeper level, I think we should consider to create a separate Log for Dry Run validations? This keeps the Kubernetes objects separate from ones in the workspace (they act in isolation).

[f1ames]

Some quick suggestions I had when thinking about this:

• From what I found it's not possible to directly read output of terminal in VSC. There are some workarounds, like https://stackoverflow.com/a/74530249/646871.
• We can provide a configuration option when user can define what command to run to generate resources (be it Helm or Kustomize) and run it on every change. Since the extension will manage when and how it is run it should be possible to capture output.

On a deeper level, I think we should consider to create a separate Log for Dry Run validations? This keeps the Kubernetes objects separate from ones in the workspace (they act in isolation).

Agree 👍