It's unusual, but very interesting!

Thank you!

I don't understand where the reference solution finds additional user authorship. Regardless of whether I use the API or look at the user's page, the result is the same in my case. Ok, although it's not completely clear how it works, but as a huge hint, exactly 1 API request is needed.

It's nice to go into the comments sometimes, thank you!

Chuck Norris wasn't developed code, code itself was developed for Chuck Norris...

• Python fixed in this fork
  (16 --> 80 out of 384 solutions invalidated)

• Ruby fixed in this fork
  (5 --> 14 out of 63 solutions invalidated)

• JavaScript edge cases added in this fork + edit due to assert.deepEqual failing to compare +0 & -0
  (0 --> 105 out of 493 solutions invalidated)

• Haskell edge cases added directly in the edit window
  (0 --> 5 out of 39 solutions invalidated)

@Just4FunCoder check the issue above, the reference solution doesn't handle tests like that one.

Testing for [0, 79] and 0:
[0, 79] should equal None,
but 0 * 79 == 0...

No issue with kata.

Description says: "Return [...] None if: [...] more than one number meet the conditions".

For your example, matching numbers can be 4, 8, 16, etc.

python: error on Fixed Tests 4:

right = ['The number is an even number', 'The number is divisible by 4']
wrong = ['The number is divisible by 3']
test.assert_equals(guess_number(right, wrong), None)

currect result: 4