list don't change

How people think to produce code like this? This perfect work and don't understandable for me.
idx and i can be removed.

That's strange but in solutions from my name have publicked difference code. This code is slightly different from the code that I have submitted.
The interpreter decided to publish ))

Attention! The var frogJump detected!
Your code running in range 0ms - 2ms. Cool!
Many thanks myjinxin2015

I did not take into account "All the elements are 1(absolute) difference from its neighbors."
I read this like "elements difference from its neighbors" (Even though the font is bold).
Key for optimization apparently in Math.abs(arr[i]-arr[i+1]) == 1.
I must don not forget that simultaneously master two languages: JS and native En.

Your code running in range 0ms - 14ms
No more knowledge what to delete or change i++ vs i+=1 or something else. It looks like a fine-tuning - guessing game.
One word "performance".

Finally it's done.. but with some nuance..
May be 6ms per test sometime reverse to 9ms per test..Who knows?. Just... for couple of test - not for all ))

Sometimes we can see this message in couple of tests: You code is too slow ;-) - Expected: 'Less than 6ms per testcase', instead got: '9ms'
It's Mr Performance. I have just deal with it (temporarily)

I very enjoy to solve this one. Many thanks.

The rows var prime=[] and prime.push(a) were used for debbug and don't nessesary in final solution.

Hi,
It's some troble with this. If return include last item we get indefined
console.log(maxZeroSequence([5,4, -9,8])) // [5,4,-9]
console.log(maxZeroSequence([5,4, -9])) // undefined

How good kata is - where there are no performance and others non-kosher things like BST, memoization and combinatorics (irony). I put my "very" in satisfaction rating (seriously). IMHO 5 kyu is too much.

As usual in performance problem I get Process was terminated. It took longer than...
Stability is a sign of perfect skill (irony)

n=length-1
indexOf(n)

Apparently indexOf(n) spent all my time ))