This study presents a memory-based sliding mode control for singular semi-Markov jump systems using a novel dynamic-memory event-triggered protocol. Based on the average dwell-time strategy, a deterministic switching signal was developed to adjust the variation of the semi-Markov chains. Randomly occurring deception attacks are considered owing to the existence of an adversary. Furthermore, to enhance transmission efficiency and achieve better control performance, a novel dynamic-memory triggering condition is proposed, in which both historical transmitted data and two auxiliary dynamic variables are implemented. Based on the proposed protocol and Lyapunov theory, some parameter-dependent sufficient criteria are established to guarantee mean-square exponential stability and strictly dissipative performance, and the desired memory-based sliding mode controller is designed. Finally, two simulation examples were presented to verify the effectiveness of the proposed methodology.