Unusable VPC Endpoint caused by empty list for subnet_ids when using non-standard Environment name

Version: 2.7.3

What have been done: Deploy a VPC Endpoint in an environment named prd2 from the Gitlab CI/CD PoV, but with subnets named -z2-prd-

What was observed: All VPC Endpoints are created with an empty subnet list, making them unusable. EC2 Instance failed to startup properly because it was not able to reach a vpc endpoint in its boot process.

What was expected: Reference subnet ids from the terraform state of the underlying terraform layer. Otherwise, raise an error or at least a warning.