For many challenging applications in simulation technology, micro-scale phenomena can dominate macro-scale behavior. Using micro-scale models to simulate the complete macro-scale domain of interest is however oftentimes infeasible due to excessive computational cost. Thus, a multiscale consideration is necessary. In this work, we present a software solution to couple a hierarchy of surrogate models on the micro-scale to a macro-scale model using the coupling library preCICE and the Micro Manager. preCICE is designed to couple two or more simulations on the same scale. The Micro Manager creates a large number of micro simulations from a user-created library, and adaptively controls them. This approach reuses coupling functionality of preCICE while also preserving the black-box coupling philosophy. The Micro Manager has two adaptivity strategies. The first strategy is taken from Redeker and Eck, 2013, in which micro simulations are marked as active or inactive based on a user-tuned similarity criteria. To get the complete micro-scale solution field, the inactive simulations are associated to their most-similar active counterparts. In the second strategy, the user provides not just one model as a micro simulation, but instead a hierarchy of surrogate models. To create these surrogate models, the Micro Manager has a snapshot creation functionality. Based on a user-defined rule-set, the Micro Manager switches from a full-order model to a reduced-order model. We show the workings of both the adaptivity strategies for a two-scale heat conduction problem in which we resolve a grain micro-structure to upscale quantities necessary on the macro scale. We especially look at scalability and efficiency of both strategies for large problems solved on HPC infrastructure. We apply our software solution for problems in the field of mechanics models of composite structures, and biomechanical models of human organs.