Troy
cc Justin
Ann
Brandon
Hello, team! I've been checking logs on daily test sessions of the Daryle AI app (main one, not extension) and I noticed an influx of Partial Passes as results.
Note that a Partial Pass doesn't mean a full-on bug, but areas that are lacking based on the parameters and set goals for the app. Usually it's missing citations, context, etc. Generally ares we can improve on for Daryle to give more robust responses.
I've compiled them here in this report which synthesized common behavior found by the QA tool. This way, we can adress them with specificity.