Not strictly related to this PR, but I wonder if we want to rename this to AttributionAggregationProtocol, as otherwise it might be misinterpreted as what the spec calls attribution logic.

Originally posted by @apasel422 in #203 (comment)