How can I set up the jump distribution for a CompoundPoissonProcess?

The definition of the CompoundPoissonProcess mentions being able to control jump sizes:

A compound Poisson process for modeling jump processes.

The process has jumps that occur according to a Poisson process with given rate,
and jump sizes determined by a specified distribution.

Still, I haven't been able to find the syntax for this in the code or documentation. There seem to be fields only for the rates, but no mention of the distribution for the jump size/direction:

# Fields
- `rate`: Jump rate function or constant (λ parameter)
- `currate`: Current rate value (cached for efficiency)
- `computerates`: Whether to recompute rates at each step

Is there a function that allows me to model the probability distribution for the jump as well? I.e. with a Distributions.jl object or some other type for which rand is well defined?

Start with src/compoundpoisson.jl and the linked CompoundPoissonProcess documentation, especially its fields and constructor details. Determine whether jump-size distributions are supported; done means the supported syntax is documented or the requested capability is clearly specified for implementation.