Can we improve on the approach we used in the SK repository
Can we improve on the approach we used in the SK repository