This PR updates the .gitignore file that is created by the myks init command to reflect the latest changes in the .myks directory.

UPD: decided to just exclude the whole .myks directory by default.

With these changes, the following files are ignored:

• .myks/tmp: data schema and values files managed by myks and refreshed on every run
• .myks/envs/**/helm: helm values files generated by myks on every render of this app
• .myks/envs/**/steps: yaml files with intermediate results of the rendering steps
• .myks/envs/**/argocd_defaults.ytt.yaml: generated on every run default data values for the ArgoCD plugin

What is not ignored by default (these files are useful for debugging and tracking upstream changes in PRs):

• .myks/vendir-cache: vendir cache entries
  • this probably should be ignored by default, users can decide whether to carry the cache in git or not;
    I personally keep it in git to track upstream changes, but now, for every update, there are a lot of renames that clutter PR diffs
    Here is an example of such PR: https://github.com/Zebradil/myks-homelab/pull/130/files
• .myks/envs/**/env-data.yaml: generated data values files for environments, useful to see the final definition of an environment
• .myks/envs/**/vendor: directory with links to the vendir cache entries
• .myks/envs/**/vendir-links.yaml: map of vendor directories to cache entries
• .myks/envs/**/vendir.yaml: generated vendir configuration file for this application

It feels like we can completely ignore the .myks directory by default and provide hints to users to include some of the ignored files and directories to satisfy particular use cases.