This paper proposes a task-aware distributed inter-layer topology optimization method for resource-limited LEO-LEO satellite networks. The method is designed to maximize observation benefits with limited onboard resources by optimizing inter-layer links between Remote-Sensing Layer Satellites (RLSs) and Communication Layer Satellites (CLSs). The approach considers time-space relationships between the two layers and incorporates practical constraints such as link switching intervals for laser transceivers. A new time-slot division method based on link switching intervals is introduced to linearize the time-sequential coupling problem. Simulation results in four constellation systems show that the proposed distributed interactive method can achieve an optimal solution close to the centralized global solution with much less time. The method supports onboard optimization for future space networks by considering task differences, link switching time, and resource management. The paper also presents a distributed interactive mechanism between RLSs and CLSs, which allows for efficient link selection and data transmission. The proposed method is validated through simulations and is shown to be effective in optimizing data transmission in LEO DLSNs. The key contributions include modeling realistic inter-layer laser connectable conditions, proposing a new time-slot division method, and developing a distributed inter-layer topology optimization method based on time-space relationships. The method is implemented in a distributed system and is shown to be efficient in solving the optimization problem with reduced computational complexity.This paper proposes a task-aware distributed inter-layer topology optimization method for resource-limited LEO-LEO satellite networks. The method is designed to maximize observation benefits with limited onboard resources by optimizing inter-layer links between Remote-Sensing Layer Satellites (RLSs) and Communication Layer Satellites (CLSs). The approach considers time-space relationships between the two layers and incorporates practical constraints such as link switching intervals for laser transceivers. A new time-slot division method based on link switching intervals is introduced to linearize the time-sequential coupling problem. Simulation results in four constellation systems show that the proposed distributed interactive method can achieve an optimal solution close to the centralized global solution with much less time. The method supports onboard optimization for future space networks by considering task differences, link switching time, and resource management. The paper also presents a distributed interactive mechanism between RLSs and CLSs, which allows for efficient link selection and data transmission. The proposed method is validated through simulations and is shown to be effective in optimizing data transmission in LEO DLSNs. The key contributions include modeling realistic inter-layer laser connectable conditions, proposing a new time-slot division method, and developing a distributed inter-layer topology optimization method based on time-space relationships. The method is implemented in a distributed system and is shown to be efficient in solving the optimization problem with reduced computational complexity.