the rule seems to check whether or not the loop has a break after the mutation, because that makes the mutation safe:

items = [1, 2, 3]

for item in items:
    items.append(item) # no error
    break

however i've noticed two issues with this logic:

1. it doesn't seem to account for when there's another statement between the mutation and the break:

   items = [1, 2, 3]
   
   for item in items:
       items.append(item)
       if True:
           pass
       break # error

   this should still be considered safe, because the statement does not effect the flow of the loop at all.

2. if there's a continue statement followed by a break statement, it does not report an error, even though the mutation is unsafe:

   items = [1, 2, 3]
   
   for item in items:
       items.remove(item)
       continue
       break

playground