Quadruped robots represent a promising platform for autonomous operations in indoor environments, thanks to their ability to navigate terrain inaccessible to wheeled platforms, such as staircases, raised thresholds, and uneven surfaces. However, the development of a complete and reliable system for managing autonomous missions on these robots remains an engineering challenge requiring the integration of heterogeneous software components. This thesis presents the design and implementation of a system for managing autonomous missions on the Unitree Go2 quadruped robot, built entirely within the ROS 2 framework. The proposed architecture is organised into two subsystems that aim to enable the robot to operate autonomously in environments spanning multiple elevation levels.