Meta-models are the centrepiece of Model Driven Engineering, required in many activities: modelling, creating DSLs (Domain Specific Languages), xDSLs (executable DSLs), or writing model transformations. Therefore, designing meta-models should be done carefully but it could be a complicated task with large ones. Meta-models should then be validated but it is mostly done considering their usability: in particular if it is possible to instantiate them. Automatic model generators are used and if they are unable to generate models it means the meta-model with its instantiation parameters (e.g. size of the models) is wrong. Several generators exist, but most of them have binary output: success or failure, without helping the meta-model debugging. In this paper, we introduce an approach, in which we statically analyse a meta-model with its instantiation parameters. In this first work, we detect inconsistencies considering each reference or each inheritance separately. Therefore we provide feedback to the meta-model designer to help her to debug the meta-model.