When the user input a value that doesn’t match the data type an error is raised and prompt to the user. In this case it looks that there is no way to use the input data, is it possible to retrieve it? I think that having the chance to retrieve the input that generated an error could be used in useful ways a script.
The utterances from the user are available in context.userInputs, but note that any particular utterance may contain valid input for other entities that are either extracted in the current volley or will be extracted in the future once the dialog reaches that point.
Therefore it is not always a simple case of echoing the entire utterance back to the user and saying that this is invalid. In conversation you cannot always assume that the user’s utterance is related solely to the previous prompt.