Concurrent Constraint Programming (CCP) is a declarative programming technique, used in scheduling, production planning, and problems of Operations Research (Management Science). The goal is to find a solution for a set of variables: a corresponding set of values that satisfy a given set of constraints (relationships). Several processes (called "constraint propagators") will be continuously watching the domains (feasible values) of the set of variables they constrain. As soon as the domain of one of their variables is shrinking, they will try to also shrink the domain of the others corresponding to the constrained relation they guard. We explain how this technique works and demo its application to some recreational puzzles: N-Queens, Sudoku, a Probabilistic Mine Sweeper, ...