I typed your query into Google ("test aws instance status checks") and found this page on the AWS forum which has two ways to do it.
An easy way to trigger a status check failure is to disable the
default NIC in the instance. For Linux instances,
sudo ifconfig eth0 down
should do the trick.
Recovery is by adding an additional ENI to the instance and connecting
in through that, then bringing the original ENI up with "sudo ifconfig
The same page has this
I know this is an old post, but I saw a suggestion from another post
It looks like you can temporarily set a different state for an alarm.
I mention temporary because during the next period, the alarm will be
re-evaluated for it's new state.
Taking a look at the docs I think you should be able to do the following:
aws cloudwatch set-alarm-state \
--alarm-name "ec2_system_alarm" \
--state-value ALARM \
--state-reason "simulate an ec2 system failure"