Thanks. I was going to point out that this was an existing problem in the Fediverse (there are instances that are 'unsafe' because they are either explicitly _for_ hate speech or just don't do enough to moderate it) and that the standard approach is to not federate with those instances, nor with any instance that chooses to federate with them. It's not universally popular (some people don't like the idea of 'guilt by federation') but it's necessary if your goal is to prevent your users from coming into contact with nazis.