I’ll start. For me, to make building skills much faster and more powerful, I’d like to have a in-skill library of the following:
- face animations (ex: weather-, time-, and action-related sequences),
- jibo lower level sounds (ex: laughing, inquisitive, robot-like, and other sounds),
- body movement animations (ex: dancing, wiggling, stretching),
- full behaviors (ex: feelings like fidgity, sad, confused, as well as common tasks like listings).
That would allow me to put together highly dynamic skills by only putting together the logic and save a lot of time. It would also create a consistency across skills.