CS50AI by Havard University

Recursive-Depth-First-Graph-Search as part of Harvardx's CS50AI :

This project is characterized by the straightforwardness and simplicity with which it was implemented. It solves Tic-Tac-Toe, basically any finite 2-player game, using a recursive depth-first graph search. The game Tic-Tac-Toe was adopted from the pygames implementation.
Since the game is finite and the strategy space limited, the AI can calculate every terminal situation and select the most advantageous outcome. By recursively climbing up the graph, which can be thought of as a classic tree, the AI determines the optimal move given any, and in particular the current state of the game-board. As a result, the AI can never lose, but you as a human adversary could lose and always will should you not play optimally.