Using Zenith with Ponytail ? + zenith-hooks

Hi there,
Do you know if ponytail should be used alongside Zenith?

the thing about Ponytail is that it is really good for code, writes less code that does more
but the caveate is that it is lazy when it comes to testing,
so if all aspects of the flow and all types of agents use it as a global enabled skill
that probably wont be so good
but specifically for code builder agents it should be great

i wonder i edit the Ponytail skill to specifically use itself only for strict code writing and building jobs, but not to enable it if running tests, or other non coding tasks
if that would be enough to make it work only when it needs to.

my hypothesis is that it will work fine if i make that change
but lets say it puts the trust in the llm to choose, rather then specific hooks, stages in a deterministic pipeline or flow
like im guessing is heavily used in zenith (some parts programatic, some agentic)
so maybe in the future, if zenith could expose its own hooks, then more programatic solutions could be created

what do you think? about ponytail + zenith, and zenith-hooks
Thanks a lot and all the best!

The issue identifies no files, tests, or entry points. First clarify whether the requested work is Ponytail integration, deterministic hooks, or both, then define acceptance criteria for when skills run and how Zenith exposes hooks.