[rules-dev] Justification based TMS