It would be great to have a bit more control over the specific run command used to trigger our dbt models. 

At the moment the command used is dbt run --models +[model_name] which at present means that we are having to make changes to our models that goes against the advised process with dbt.

As we have a number of intermediate models with some variation in the materialisation conditions used in the models that make up the reporting models that we have scheduled as transformations in Fivetran, we have noticed in particular that the "+" operator used before our model name often results in models that we reference in our reporting models running on a schedule that is unsuitable for our requirements.

This essentially means that where we have large models which we would like to run on their own schedule independent from their endpoint model being referenced via {{ ref("name")) }}, which does resolve to the full production model name within the compiled query but still runs the actual query for those models on the same schedule as the endpoint model.

Ideally it would be fantastic to have some degree of control over this, potentially by providing a flag when scheduling a transformation in Fivetran to state if we would like to also run the downstream/upstream models.