.

Old kata syndrome, re-ranking them is not possible now due to the excessive amount of solutions that is associated with legacy code in DB, but it will be addressed soon as admins and the CW team are aware of this ^^

some mistakes that i see:

• you allocate size bytes, but then you access ret[size], this is an out-of-bounds access, the last character is at ret[size - 1]
• ceil(n / 2) is equivalent to just n / 2, since you are diving 2 integers
• the diamond must have a trailing newline (\n), but you are not adding one

• r, g and b can be negative. The format "%x" takes an unsigned integer as input, so you are printing a very large number after reinterpretation of the signed int as un unsigned int. Since the output string has enough room for 6 characters + 1 nul terminator, you are writing out-of-bounds, which causes chaos in the tests suit
• sprintf(h, "%c", '\0'); appends 2 nul characters to the string, because sprintf() always nul terminates its target string after finishing the printing. same problem as above, you are writing out-of-bounds

See the initial code in C:

void rgb (int r, int g, int b, char hex[6 + 1])
{
  // write to the hex string
  *hex = '\0';
}

That's a problem with your code, not a kata issue, use Question label next time.