An operations team wants to automate the remediation of recurring EC2 configuration drift by running a repeatable, predefined set of steps invoked from a script. Which approach aligns with the exam's remediation objectives?