Perhaps retrace your steps as to discern at which point you were mistaken in perceiving that the kata had some functional flaw. Take note that the JavaScript version of this kata has been solved over 1K times and please keep in mind that in order to post an issue on a kata, you should be able to explain how and why it occurs, and provide examples of test cases where the breakdown occurs, such as: input, expected output, your code result, and the like. Otherwise, you should post as a question instead. Nevertheless, given that you no longer perceive any flaw, you should mark this as resolved.

Yeah but just like in the movie, it would not help too much. Thousands of eggs were laid, hundreds of them already hatched, and Ripley still needed her flamethrower to deal with the rest.

Let'just hope that nuking the place from orbit is not the only option :D

Probably you edited the sample tests. Not a kata issue.

const { assert } = require('chai');

describe("Split Strings", () => {
  it("Basic tests", () => {
    assert.deepEqual(solution("abcdef"), ["ab", "cd", "ef"]);
    assert.deepEqual(solution("abcdefg"), ["ab", "cd", "ef", "g_"]);
    assert.deepEqual(solution(""), []);
  });
});

These are the current sample tests. There is no input "C" there.

Probably when those solutions passed the tests, the tests were more lenient. Automatic revalidationg doesn't trigger for katas with many solutions, so it has to be done manually.

Not anymore.