Duplicate database leaves orphan archivelog files

I am cloning the production database to another one, lets say CLONEDB, in a regular basis. I am using rman duplicate database script like:

run {
duplicate target database to CLONEDB
FROM ACTIVE DATABASE
PASSWORD FILE
SPFILE
SET ....
DB_FILE_NAME_CONVERT=....
LOGFILE
  GROUP 1 (
     ....
  ) SIZE 100M,
....
;
}

Before running the script I am doing an archivelog rman backup of the CLONEDB (it deletes backed up archivelogs) and DROP DATABASE CLONEDB.

While running duplicate database, rman copies datafiles and archivelogs from the source database. For example, I can see that such file has been copied:

input archived log thread=1 sequence=666 ...
output file name=+DYSK/archivelog/2015_01_16/thread_1_seq_666
channel ORA_DISK_1: archived log copy complete, elapsed time: 00:00:01

Then rman recovers database (via set until scn 666666; recover clone database delete archivelog;), so the archivelog above is applied to the database:

starting media recovery
archived log for thread 1 with sequence 666 is already on disk as file +DYSK/archivelog/2015_01_16/thread_1_seq_666
archived log file name=+DYSK/archivelog/2015_01_16/thread_1_seq_666 thread=1 sequence=666
media recovery complete, elapsed time: 00:00:02
Finished recover at 2015.01.16 16:21:57
Oracle instance started

Everything is OK, database starts up, users are happy. But the archivelog file (in this example thread_1_seq_666) stays on the disk. And because rman changes the database incarnation after duplicate process, this file is never deleted during the backup process, as it belogns to the previous incarnation.

So from time to time I must do the housekeeping and delete old, orphaned archivelogs directly from the disk. Of course I would like to do it automatically, with the duplication script. I could add rm command from asmcmd for the old files, but maybe there is a way to convince rman to do it for me during the duplicate database?