"Vibe Coding"

So, question is,

How / what ontology should be defined, to better describe 'vibe coding'...

I've been experimenting - fwiw ; it's been great!  generally...

my experience has been, that no single model can 'do the job', mostly, they
do parts, and you've got to use different 'models' (llms, etc.) to get the
'job' done.

also depends on how complex the 'job' is...  but, then you get to tools
like respec, where its important to honourably define - who did what, and
if users are 'vibe coding', then, they're not actually the primary
'software developer'... rather, they're providing instructions...

a bit like, defining the text in a microsoft word document; rather than,
defining the code of a microsoft word document....  (or indeed also, the
application?)

So,

whilst i look forward to providing an update about WIP (works in progress);
i thought, i'd prompt the list, with the query - what do you think, how
should the definitions of 'who did what' apply to AI agents, particularly
when there's alot of them involved; and, how could this in-turn be defined
as some sort of respec doc, to define the characteristics of what might be
defined as 'best practice'..

I suspect it may intersect with git, and other elements; but, I thought I'd
prompt others to provide their view, in case that's going to give results!

fwiw; responses on the list are preferred...  at least, for me..  gives a
historical view of how things happened overtime.  i've lost alot of the
records from skype or other platforms, some most involved in the most
important works, now deceased...   anyhow.  i also understand the
importance of being able to reach-out, people are welcome to do so... but,
as noted, crafting considerations into something that can be put onto the
list - totally preferred (where appropriate), for me 🙏

Tim.h.