Microgrids are increasingly operated in a decentralized manner. However, coordinating interconnected microgrids with heterogeneous energy profiles, local physical constraints, and distributed storage remains challenging for achieving collective energy efficiency and self-consumption. This paper proposes a decentralized peer-to-peer energy trading model for interconnected microgrids, in which trading decisions are driven solely by the local energy context of each microgrid and its physical constraints. The proposed approach enables coordinated balancing of supply and demand at local network, reducing reliance on the main grid. Microgrids with an energy surplus act as sellers and adjust their energy bidding behavior using aggressive–conservative strategies. Their selection strategy is linked to available energy and state of the local energy storage. Microgrids experiencing an energy deficit act as buyers, with predefined energy requirements and no strategic decision-making, as energy transactions occur at a fixed internal price within the network. A coordination mechanism was defined to match buyers and sellers and to determine feasible energy exchange volumes under energy network physical constraints. A genetic algorithm is employed to identify compatible microgrid pairs and optimal energy transfers, with the objective of maximizing network-level energy stability while preventing unsustainable operating conditions. The model was validated on a simulated network of 100 heterogeneous microgrids, each with varying energy storage capacity and stability thresholds. The results indicate that 96 microgrids reached a stable energy state after peer-to-peer trading, demonstrating the effectiveness of the proposed coordination mechanism in identifying compatible energy exchange pairs and significantly improving network-wide energy stability.