Would leave this here for discussion & as possible future plan.
Be able to add transforms in semi-private api. They should be invoked with
- markup
- each slide markup
- each slide html
- resulting html
And returned values should take place of inital values (modified markup/html). Need to be sure to account for possibility to return multiple slides from one or none.