Great kata to practice some performance optimization on the fundamentals.

Please allow results with or without initial (0,0).

Outputs:

A list of tuples, with the state of the two jugs after every step

That initial [0,0] should not be there, and that is specified ( note it says "after" every step ).

Not a kata issue - but I am going to raise it as a suggestion; that initial (0,0) can very easily be checked for and passed.

( JS )

Function.prototype.toString should be nuked altogether in Preloaded.

Just setting _checker.toString to something leaves the door open for Function.prototype.toString.call(_checker), which still reads the source code.

We need to see your code to know why it fails.

Hello. I faced a problem while solving a JS task: tests [3, 5, 4, "It is possible to reach n=4 liters in one jug"] fail. The correct answer [3, 4] is specified in the task description, I even tried to substitute it manually into the return, but it didn't work. It looks like there is an issue in the _checker function in the tests. The test fails with this error:
TypeError: l.at(...).includes is not a function
at _checker (test.js:17:17)
at act (test.js:177:19)
at Context. (test.js:166:13)
at process.processImmediate (node:internal/timers:471:21)

Translated with DeepL.com (free version)

fixed

approved to avoid merge conflicts later on

this should be fixed before approving the JS translation

So unfortunately there is an easy way to break this and succeed at all of the tests:
def wpp(a, b, n):
if n > b and n > a:
return []
if (not a % 2 and not b % 2 and n % 2) or ((not a % b or not b % a) and not n % 2 and n < a and n < b):
return []
return [(n, min(a, b))]

JS Translation

Thanks. ;)

Fixed your assertions. This issue is now solved.

I implemented a fix. Your solution now passes.