All languages should have the edge case of n = 1 added (Certain languages have built-in range functions with a step parameter and will not work when step > the end parameter)

Lua translation!

In C teams is already initialized with zeros in all tests. Since it's not documented, there should be tests where it's not initialized.

IMHO current constraints makes it 6kyu.
At 7kyu brute-force solutions should be accepted.

Groovy Translation

You need to test for the case when n is the divisor and 2n+1 is the bound and 2n is the solution. I passed the test using a for loop but I would have failed this test so I got a false positive.

Relational databases assume that the order of rows in the resulting set doesn't matter, BUT in this exercise different rows order leads to unacceptable solution. It looks like we need a specific order clause to solve this misconception

One random test Random tests
expected 'Not Balanced' to equal 'Balanced' is not passing. Idk what to do. please someone look into my code.

Whats the take on using Itertools in rust? it's not part of the rust native library, but a lot of solutions are implementing it to make their solutions more functional. Is it okay to use external crates or should we rely only on the native library?

Below Powershell 7.2 answer should work. Returns the correct answer locally, but returns 0 on here.

function maxTriSum ([int[]]$numbers)
{
  [int[]]$intArray = @()
  
  $numbers = $numbers | sort -descending
  
  for($i=0;$i -lt $numbers.length -and $intArray.length -lt 3;$i++)
  {
    if($numbers[$i] -ne $numbers[$i -1])
    {
      $intArray += $numbers[$i]
    }
  }
  
  return ($intArray | Measure-Object -sum).sum
}

This has a performance tag, but O(n^2) solutions pass in Python. Is that intended?