"Should return null for invalid arguments" The last check yields a number, but should return null. I used the isNaN method, but it doesn't work. Are you using standard primitive data in your test?

My solutions fails in 3 tests when radius, arc and angle are given and instead of null, in 2 tests solution gives false and in 1 - instead of true gives false

This one definitely was hiding under the radar.
An easy six, although the input validation is a nuisance.

Approved by someone

Are you sure those are actually numbers?

Julia translation

Added limited random tests ( expected value true will be generated exceedingly rarely. relying on fixed tests for that ).

Smile, you missed an edge condition where you're returning NaN instead of null. You're not going to fix that anymore, I know. :P

And approved.

If you're going to write a wrapper function for testing, make sure it can't be hijacked by solver ( defining it with const is one way to do that ).

Native Math does not have enumerable properties. In keeping with this, and to teach good habits, defined methods should not be enumerable, and this should be tested for.

toFixed does rounding differently from Math.round ( and weirdly in general ). This behaviour should be warned for and toFixed should ideally be discouraged and tested against.

Rounding is not the appropriate solution for floating point representation inaccuracy. This has been dissed as an issue below, but again, it's teaching bad habits.

Not an issue

Updated description.

This is fixed, though not ideal.

What kind of rounding is expected for values ending with 5?
Existing solutions return different results for Math.toRadians(-60.87676573264997) and they all pass the tests.

All tests are valid integers

No really integers...