Software architects have been recently adopting self-organizing strategies to design dynamic and adaptive decentralized systems capable of executing without interference from the user, from the administrator or from other systems. This self-organization process usually results in unpredicted emergent behaviors resulting from the interaction between the agents. Emergent behaviors can be beneficial or harmful to the system execution and organization. Because of this, multiagent systems will only be adopted into domains that involve risk - like industries, hospitals, military equipments, etc. - if one can guarantee that it will be able to fulfill its design goals. However, it is not easy to guarantee this, since it is impossible to map all the possible behaviors of a system modeled through agent interaction. Some approaches have already been proposed, but their use is usually extremely complex. Therefore, this work presents a new method for the experimental verification of self-organizing multiagent systems, evolving the empiric verification technique proposed by Kevrekidis and complementing it through an autonomic approach linked to the simplicity of the verification through online planners. As a result, a framework for designing, verifying and simulating multiagent systems is presented.