There are several methods in klampt.plan.motionplanning which admittedly are not particularly well documented due to some SWIG formatting issues. You can look at the C documentation in Klampt/Python/klampt/src/interpolate.h for some help.

If you would like to do a minimum-time path between joint space milestones (where you don't need to interpolate the trajectory exactly) you can use interpolate_nd_min_time. If you would like to interpolate the straight line joint space path exactly, you can use interpolate_nd_min_time_linear

A more Pythonic way of getting access to this is klampt.control.motion_generation.AccelerationBoundedMotionGeneration. You can set the start state, call setTarget multiple times, and then call trajectory to receive the trajectory.