This appears to be a python program that uses the vulnerable MD5 hash to protect a flag.

The python program provides output of g, p, and A when run, then asks for (numeric) user input of k which is a shared secret. It then gives cyphertext based on the k value given (p and A are different each time though g appears to be constant) Each time it is run it appears to generate a new shared secret

My current thinking is that hashcat can be used on the initial output to find the shared secret which can then be entered as k to output the flag as c.