Cooperative control is concerned with engineering systems that can be characterized as a collection of decision making components called agents with locally sensed information and limited inter-component communication, all seeking to achieve a collective objective. Examples include mobile sensor networks, autonomous vehicle systems, distributed computation, and power systems. Areas of research that are related to cooperative control include multi-agent control, distributed systems, networked control, etc. Most of the results in cooperative control (such as synchronization) of MAS require some prior information of the communication topology, which means that control or synchronization might not be achieved when there is a change in the communication topology. In particular, scale fragility has shown that instability of the network often occurs when the number of agents grows large. In this paper, we focus on scalable synchronization of MAS in the presence of disturbances and measurement noise with known frequencies. We develop a scale-free collaborative protocol design for this class of MAS. The scalable exact output and regulated output synchronizations are developed for homogeneous MAS with non-introspective agents and heterogeneous MAS with introspective agents, respectively. The proposed protocol design methodology is solely based on agent models and does not need any information about the communication network or the number of agents. The results are illustrated in numerical examples with the random number of agents.