Okay, thanks for the honest feedback, I will try again with a new concept!

I think overall the kata is rather fine as far as execution goes, it is just that the concept and idea behind it isn't new or unique, and overall isn't super interesting. I get that it is for beginners, but basic conditions, and string splitting and filtering has already been done a lot.

So my advice would be to really think of a fun, unique idea, and try again. (Often that means aiming for a bit of a harder kata). I doubt this one will make it through beta.

Hi, I would appreciate any feedback to improve the kata. Thanks!

Not random enough, IMHO, you could construct a string with a random number of random words (not taken from a dict or a list).
The string or even the words do not need to make sense.
Just my opinion though, I wont reraise the issue.

I've made some random tests now

I think it is important to design those tests carefully before publishing the kata, to avoid downvoting.

A little oversight by myself there, thanks for the feedback!

Missing random test cases.
See https://docs.codewars.com/authoring/guidelines/submission-tests/