Yes, you're right. Fixed it.

Yes, my fault, sorry. I really misunderstood the test.
There are no problems with kata, thanks for the explanation.

Then it turns out that:
"It should work for random inputs too: 6030 should equal 6003"
The initial value is: "6003"?
The answer is: "6030"?
Then my code works correctly and everything corresponds to the correct answers, but the kata does not pass.

Initial values:
59884848495583;
6030;
640010;
917.

obviously, this is a value that should be less than the initial one

This is in the sense that the answer to the test already contradicts the task. The answer is less than the original number.

I worked on the bugs and fixed everything.

You correctly noticed that another language is not a reason to create another kata. But I also wrote there that we have different conditions of the task, the level of the task and different requirements for the solution.

My home compiler in VS code did not issue this warning, but I will listen to you and make edits.

Changes have been made, now the tests include random tests.

I'll fix it.

Another user's task talks about writing a cryptographer and a decryptor, while I only need to write a cryptographer, there is also no need to check for a mutually prime number, the provided gcd function does it there, in my task the user must organize a check for a mutually prime number himself.
I also write in C++

My fault. I'll make changes.

Mutually prime numbers are integers that have no common divisors other than ±1