Wheeled-legged robots have the potential to navigate in cluttered and irregular scenarios by altering the locomotion modes to adapt to the terrain challenges and effectively reach targeted locations in unstructured spaces. To achieve this functionality, a hybrid locomotion planner is necessary. In this work we present a search-based planner, which explores a set of motion primitives and a 2.5D traversability map extracted from the environment to generate navigation plans for the hybrid mobility robot CENTAURO. The planner explores the map from the current robot position to the goal location requested by the user, considering the most appropriate composition and tuning of locomotion primitives to build up a feasible plan, which is then executed by the robot. The available primitives are prioritized and can be easily modified, added or removed through a configuration file. Our approach was evaluated both in simulation and on the real wheeled-legged robot CENTAURO, demonstrating traversing capabilities in cluttered environments with various obstacles.