I did go ahead and test this. I didn't do a proper "lab conditions" style test, but I'm reasonably satisfied that ripostes aren't proccing defensive procs. A bard with no weapons singing Reverberating Barrier, beating on a /dueled rogue with an empty mainhand, did not result in stun procs hitting the rogue while standing there idly, or while using the Counterattack disc, whitch makes every swing a riposte for a couple ticks. On the flipside, the rogue hitting the bard with regular autoattack did proc them, about 8 times in the short time I was testing that. That and the riposte version being a zero-variance case, hits statistical confidence reasonably quickly. The part where it wasn't "proper conditions", is I probably should've test copied, frozen buffs and done 2-3 minutes at least of straight autoattacking on the Counterattack disc looking for a defensive proc, then waited long enough to time out the fight before re-testing the rogue's autoattack to see it proc successfully off of that. Instead I had to look at the parse log to make sure all the defensive procs happened during a timeframe where the piercing weapon in my rogue's offhand was firing. Aside from that, there is some possibility NPC and PC are for some reason handled differently on this. I saw no evidence they're different when I tried it with a pet, but I also couldn't get anywhere near as many ripostes in a short timeframe, so It'd need to be a much longer test. Maybe someday I'll get a few people to help me set up something with roots, heals, stuns, and a Sarnak's auto-riposte ability to manipulate a somewhat quick version of an NPC test.