Recently, biological data has been increasingly produced calling for the existence of computational models able to organize and computationally reproduce existing observations. In particular, biological regulatory networks have been modeled relying on the Sign Consistency Model or the logical formalism. However, their construction still completely relies on a domain expert to choose the best functions for every network component. Due to the number of possible functions for k arguments, this is typically a process prone to error. Here, we propose to assist the modeler using logic-based tools to verify the model, identifying crucial network components responsible for model inconsistency. We intend to obtain a model building procedure capable of providing the modeler with repaired models satisfying a set of pre-defined criteria, therefore minimizing possible modeling errors.