I read the source and it looks like this tool is using some kind of lattice attack[1][2] against weak k nonces that aren't "uniformly generated at random" (to quote the paper below)
You won't get great performance out of this though, because it's a CPU implementation with no GPU acceleration. Even the abstract of
Breitner and Heninger (2019) says they only computed hundreds of bitcoin private keys with this software so obviously the speed is not great (in the millions/second range that we are all used to).
[1]:
https://en.wikipedia.org/wiki/Lattice-based_cryptography[2]:
https://crypto.stackexchange.com/questions/26547/how-can-a-lattice-attack-be-applied-to-ecdsa-signatures