Everything you are referring to is the CI/CD pipeline. GitHub actions, Gitlab Runners, ArgoCD; they can all do some sort of gitops. Those dependencies existed before gitops anyway, so nothing new is being added.
You can disable reconciliation and then manually apply a change if your git repo is unavailable.
You mean like storing source code? CICD pipeline definitions? Container image specs? IaC manifests? Service configuration files?
I don't think that arbitrarily drawing the line on what package version is being deployed is a rational opinion to have. I mean, look at npm's package.json. Is it ok to specify which dependency versions you bundle with your software, but specifying which version you're deploying is unheard of?
The only drawback I see in GitOps is that CICD pipelines are still lagging way behind on providing adequate visualization strategies to monitor the current state of multi-environment deployments.
Even without knowing what you mean by git plant. I can tell you you're holding it wrong.
I'd like to help, but I can't offer useful suggestions if I have no idea what you're even doing, or prohibiting.
Fortunately, the whole thing is git based, so you have all the tools you need to do it.
What are you talking about. And whatever it is, it’s wrong.