Spawned from an email discussion regarding Tekton's impact on cluster stability should tasks be allowed to repeatedly execute unbounded.

https://tekton.dev/docs/operator/tektonconfig/#pruner

Tekton has an auto pruning feature that could potentially disrupt our ability to debug resources post-migration.

Let's evaluate potential issues here, and explore avenues to address. One option that comes to mind is that ideally you'd have some granular control in the pruner to be able to configure it to prune some objects, but not others. Perhaps via label selector include/exclude lists.

Is this something that could be contributed upstream? Should we present our use-case to the community?

CC: shawnhurley