In this paper, we consider the problem of robustifying a class of closed-loop guidance algorithms for planetary landing. Generally, such algorithms are extremely important during the terminal powered descent phase as they are critically responsible for guiding the spacecraft to the desired location with high degree of accuracy. More specifically, we explicitly describe how sliding control theory can be employed to generate energy-optimal feedback trajectories that are robust against perturbing accelerations with a known upper bound. Indeed, we show that a properly defined sliding surface can yield an acceleration command comprising a) an energy-optimal component and b) a robust component that counteracts the effect of the perturbing accelerations. Since the acceleration command is function of time-to-go, the resulting algorithm has a very peculiar behavior, where the sliding surface moves in time during the descent phase and it is in a continuous reaching mode. Its dynamics critically affect the performance of the algorithm in terms of accuracy and fuel efficient especially in off-nominal conditions. A theoretical analysis via Lyapunov stability theory shows that such class of guidance algorithms are globally finite-time stable. Simulations show that the time-dependent sliding augmentation yields superior performances versus the non-sliding counterpart. Conversely, two alternative possible formulations of the OSG yield identical results.