different behaviour for --stop between pt-kill and pt-heartbeat?

Hi,

it seems that pt-kill and pt-heartbeat behave differently in regards to the --stop and --sentinel options.

I can easily put something like this into a cron for pt-heartbeat to restart it regularly:
30 * * * * root /usr/bin/pt-heartbeat --utc --update --interval 1 --host XXX --database percona --table heartbeat --stop --sentinel /tmp/pt-heartbeat-hourly>/dev/null

I couldn’t find a way to easily achieve the same for pt-kill.

Am I missing something or is there a difference between these two tools?

Thanks
Alex