In Model-Driven Engineering, the potential advantages of using bidirectional transformations in various scenarios are largely recognized; as for instance, assuring the overall consistency of a set of interrelated models which requires the capability of propagating changes back and forth the transformation chain. Among the existing approaches, JTL (Janus Transformation Language) is a constraint-based bidirectional transformation language specifically tailored to support change propagation and non-deterministic transformations. In fact, its relational and constraint-based semantics allows to restore consistency by returning all admissible models. This paper introduces the new implementation of the language and presents the tools and its features by means of a running example.