SharpNEAT provides a full implementation of the NEAT evolutionary algorithm for evolving neural networks. It evolves both the network topology and connection weights through mutation, recombination, and selection, contrasting with fixed-structure optimization methods like backpropagation. The framework is modular by design, allowing experimentation with alternative genetic encodings and evolutionary algorithms, and includes example tasks such as biped/quadruped locomotion control, rocket flight stabilization, and digital logic design.