[seqfan] Largest k such that n is a perfect k-th power.

[Allan Wechsler]

This sequence is not defined for n=1, since 1 is a perfect k-th power for
all k. But with offset 2, we get:

1,1,2,1,1,1,3,2,1,1,1,1,1,1,2,...

which is not currently in OEIS. Now, it's a very boring sequence -- by far
most of the entries are 1. But I was surprised that this simple idea wasn't
already collected. Which is why I'm asking if there's a good reason for
this (like, my data is wrong? or an easy transform is already collected?)
instead of just creating a new entry. Any insights, sequence fans?