Gitkube VS tilt

From a workflow perspective, developers push their code to a code repository and then manually trigger a pipeline with tools such as Gitkube or Spinnaker that deploys the code in a Kubernetes environment, usually in a remote cluster in the cloud.

Could improve that crappy feedback loop :)

If the language runtimes are compiled you can't do this, but if not, in theory you shouldn't need such a stupidly long core development feedback loop.

I'm a huge fan of https://tilt.dev/ and the possibilities it unlocks for that pre-commit development.

Something like https://tilt.dev/ where you spin up a subset of the service graph in a cloud environment that hot-reloads based on local edits.

To get a similar experience of preevy up, first we’ll need to split the build and deploy using process or alternatively employ tools that orchestrate build-tag-push-update-sync flow like Skaffold/Tilt.

It's not a direct competitor, but we use https://tilt.dev/ at my company for local and remote development.

For local testing you use tilt that runs stateful services locally in a kind k8s cluster. That same config can deploy to a remote k8s server to easily share a preview of new features, which is useful for prototyping things that might not necessarily ever be merged.

There are dedicated tools just for that. Apart from skaffold check also tilt.dev, garden.io, devspace.sh, okteto.com

You basically start by downloading kind, then tilt. Then create a kind cluster with the provided configuration in the tilt repo. Then run tilt up and that's it. You'll have a fully functional Kubernetes cluster and project running complete with deployments and services. Nothing too fancy, no RBAC, no network policies etc.. Just the bare minimum to get you up and running.

I recommend also looking into tilt.

Second, and perhaps the best of all is, that I created a tilt repository located here.

But apart from the other "k8s in a box" options (like minikube, k0s, ...) you could also have a look at tilt (https://tilt.dev/), it sounds like this might be a good fit for your use case as well.