flexibility of containerization libraries

[wibed]

it might be odd to ask but the containerization repo seems like a group of libraries.
instead it feels like separate there would be more value to be had.

keeping track and reviewing all changes from one repo might be daunting if there are so many parallel issues to track.
aswell the background of deduplicating crossing similar domains such as filesystem helpers, oci registries/client...

why are these different products not kept separate? changes would be faster and better directed at the purpose of each library.
even better, it would create a landscape of packages that would be easier to overlook.

for example there might be interest for a implementation of a generator, transformer and validator to generate yaml and together with a oci client to manage local representations of charts (yes i am referring to helm here...) and deployment management of the same to k8s cluster.

why is the development of such tooling inhibited by using one collection of libraries that may create big dependencies instead of independent smaller ones that feel flexible?