Conformal Inference for XGBoost Classification using tidymodels

I really need to run conformal prediction models on classification data using XGBoost. I have seen this implemented already with tidymodels for regression tasks using the probably package, but curious to know if this can be done for classification tasks in tidymodels (or elsewhere)?

If there are any github repos with this available I'm happy to use work in progress.

If this is something that isn't on the horizon or can't be done, that's fine. Would be good to know if I need to find other solutions (or if this is something Posit are interested in developing).