This note presents a discrete-time distributed fault diagnosis and accommodation scheme for homogeneous teams of cooperative autonomous agents. Each agent estimates, via a local observer, the overall state of the team; such an estimate is used to compute both the control input to the agent itself and a set of residual vectors, sensitive to faults occurring on anyone of the teammates. The occurrence of a fault in an agent is recognized by each of the teammates, even if they are not in direct communication with each other, when the corresponding residual exceeds a suitable adaptive threshold. A recovery strategy, based on the estimate of the maximum detection time, is then applied to remove the faulty agent and to rearrange the mission. The approach is validated via numerical simulations.