Hi! I'm trying to understand the latest best practices by looking at the Fully Featured Project Example. Looking at the code, I cannot understand how dagster knows that the dbt model

story_daily_stats

depends on the

stories

asset.

story_daily_stats.sql

refers to

source('core', 'stories')

declared in sources.yml. Dagster manages to link this source table with the

stories

asset. I could be wrong but is it done implicitly by using the same asset key on both sides? • The original

stories

asset is keyed:

snowflake.core.stories

•

dagster-dbt

seems to set the asset key to

schema.model_name

with the ability to set

key_prefix

but I can't find where in this project the prefix would be set to

snowflake

🤷

Hey @Remi Gabillet you’re correct that the dependencies are determined by looking at the asset keys. You can see on this line that the dbt assets are all given the key prefix

snowflake

so that key prefix will be pre-pended to all of the sources defined in the sources.yml. Those sources are then matched to the assets with the corresponding asset keys. if you haven’t seen it yet, our dbt tutorial does a good job explaining this and how to do this for your own dbt project

Thanks you @jamie. This all makes sense now. I was close but I missed that line!