How does Prisma handle schema drift due to unsupported features such as data types, views and triggers?

You can add your views and triggers in the generated

.sql

files and they will be applied on the migration. Unsupported features can be added but you can only use raw queries and cannot fetch the value of the field.