I cannot reproduce it, at least not with this particular filter (I remember getting it fixed back in January, so it's odd that it happened in August).

That said, it ban probably happen with other Abuse filters, so we may want to handle it in some particular way, for example by checking that there's a relevant field in JSON.

I wouldn't prioritize it highly, however - it's not beautiful, but it's good enough for debugging, isn't it?..