Should build on the findings from the cross-cut spike and result in a ready-to-use mechanism (ex: a Python interceptor) that can be applied to the rest of HBI.