We ran our toolkit on four different networks of increasing complexity, starting with the absurdly simple case, and adding traps and tricks in later networks.