Haskell, 61 bytes

import Data.Numbers.Primes
until=<<((==)=<<)$sum.primeFactors

Try it online!

Explanation

until=<<((==)=<<) takes a function f and applies it to input x until a fix point is reached, that is f x equals x. primeFactors returns the list of prime factors of a number, sum yields the sum of a list of numbers.

But wait, why does until=<<((==)=<<) the job and looks so weird?

If we assume f=sum.primeFactors, a more natural definition would be until(\x->f x==x)f, because until takes a predicate (a function which returns a boolean), a function which has the same input and return type (e.g. Int -> Int) and value of this type, and then applies the function to the value until the predicate is fulfilled.

until(\x->f x==x)f is the same as until(\x->(==)(f x)x)f, and as it holds that g (h x) x is the same as (g=<<h)x, we get until(\x->((==)=<<f)x)f. After Eta conversion, this becomes until((==)=<<f)f. But if we now treat (==)=<< as a function which is applied to f, we can see that until(((==)=<<)f)f is again of the form g (h x) x, with g=until, h=((==)=<<) and x=f, so it can be rewritten to (until=<<((==)=<<))f. Using the $ operator to get rid of the outer parentheses and substituting f with sum.primeFactors yields the solution from above.

Haskell, 61 bytes

import Data.Numbers.Primes
until=<<((==)=<<)$sum.primeFactors

Try it online!

Explanation

until=<<((==)=<<) takes a function f and applies it to input x until a fix point is reached, that is f x equals x. primeFactors returns the list of prime factors of a number, sum yields the sum of a list of numbers.

But wait, why does until=<<((==)=<<) the job and looks so weird?

If we assume f=sum.primeFactors, a more natural definition would be until(\x->f x==x)f, because until takes a predicate (a function which returns a boolean), a function which has the same input and return type (e.g. Int -> Int) and value of this type, and then applies the function to the value until the predicate is fulfilled.

until(\x->f x==x)f is the same as until(\x->(==)(f x)x)f, and as it holds that g (h x) x is the same as (g=<<h)x, we get until(\x->((==)=<<f)x)f. After Eta conversion, this becomes until((==)=<<f)f. But if we now treat (==)=<< as a function which is applied to f, we can see that until(((==)=<<)f)f is again of the form g (h x) x, with g=until, h=((==)=<<) and x=f, so it can be rewritten to (until=<<((==)=<<))f. Using the $ operator to get rid of the outer parentheses and substituting f with sum.primeFactors yields the solution from above.

Haskell, 61 bytes

import Data.Numbers.Primes
until=<<((==)=<<)$sum.primeFactors

Try it online!

Haskell, 61 bytes

import Data.Numbers.Primes
until=<<((==)=<<)$sum.primeFactors

Try it online!

Explanation

until=<<((==)=<<) takes a function f and applies it to input x until a fix point is reached, that is f x equals x. primeFactors returns the list of prime factors of a number, sum yields the sum of a list of numbers.

But wait, why does until=<<((==)=<<) the job and looks so weird?

If we assume f=sum.primeFactors, a more natural definition would be until(\x->f x==x)f, because until takes a predicate (a function which returns a boolean), a function which has the same input and return type (e.g. Int -> Int) and value of this type, and then applies the function to the value until the predicate is fulfilled.

until(\x->f x==x)f is the same as until(\x->(==)(f x)x)f, and as it holds that g (h x) x is the same as (g=<<h)x, we get until(\x->((==)=<<f)x)f. After Eta conversion, this becomes until((==)=<<f)f. But if we now treat (==)=<< as a function which is applied to f, we can see that until(((==)=<<)f)f is again of the form g (h x) x, with g=until, h=((==)=<<) and x=f, so it can be rewritten to (until=<<((==)=<<))f. Using the $ operator to get rid of the outer parentheses and substituting f with sum.primeFactors yields the solution from above.