Summary In this paper, an energy management strategy of direct current (DC) microgrids (MGs) is presented. The proposed MG consists of fuel cells, batteries, and supercapacitors, along with associated DC/DC and DC/AC converters. Coordination between different sources of MG is implemented using a multi‐agent system. Multi‐agent algorithm is implemented using an open source agent building toolkit, Java Agent Development framework (JADE). The energy source components are modeled and implemented using MatLab/Simulink (power system library). The proposed energy management strategy is designed using proportional‐integral (PI) controllers, which is implemented using the JADE. Interface between MatLab/Simulink and multi‐agent system (JADE) is done with the help of MACSimJX. The proposed multi‐agent framework is presented and the interface between the JADE and MatLab/Simulink is described in details. Then, the design and implementation steps of the PI controller using JADE are presented. Simulation work is carried out, and the results show that the proposed multi‐agent‐based controller effectively coordinated with variable loads in MG.