View and manage information for existing dbt Projects¶

This topic covers how to explore the structure and metadata of an existing dbt project object. This includes viewing the project’s DAG, inspecting model and source details, and executing dbt project objects.

Browse the project DAG to see model lineage and dependencies¶

The Directed Acyclic Graph (DAG) shows how dbt models depend on each other, visualizing data lineage so you can:

Verify where a model is built (database.schema), how it materializes, and which upstream and downstream dependencies it has.
Spot and improve inefficient model designs to support better performance and scalability.

To browse the project DAG in Snowsight, navigate to Databases » your database » your schema » dbt Projects and select your project. The project details page displays the Graph of your models and their relationships, along with a Description of your project, the dbt Project definition, and Privileges.

In the Graph, click a model node to inspect model, source, or test details (such as compiled SQL and configuration) directly from the DAG.

Tip

If you’re working in a workspace, you can also reach the project details page by selecting Connect » View project from the workspace editor. For more information, see Workspaces for dbt Projects on Snowflake.

Search the DAG¶

Use the search bar in the DAG to find a model by name. Selecting a model will anchor that model as the focal point of the graph. To change the anchor node, use the search bar to search for a different model.

Control DAG depth¶

Use the upstream and downstream depth controls to change how many levels of the graph are visible relative to the anchor node. When you first anchor a model, the DAG defaults to showing 2 levels upstream and 2 levels downstream. Depth settings carry over when you anchor a different model.

The DAG displays up to 300 models at a time. If your project has more models, use search and depth controls to navigate to the part of the graph you need.

Inspect model details from the DAG¶

When you select a model node in the DAG, a side panel opens, showing:

The model’s type, file path, and a link to the target object in the database.
The row count and column count, with a list of column names.
A description of the model (if one is defined in the dbt project).
Model lineage, listing upstream and downstream dependencies with links to navigate between them.
The source and compiled SQL for the model.

On the Graph tab of the project details page, select View in Details in the side panel to open the complete model details in the object details page. From the model details view, select Show in graph to return to the Graph tab with that model anchored as the focal point.

Tip

The query history DAG and the workspace DAG share the same search, depth controls, and column-level lineage described on this page, and additionally inject runtime data (start time, end time, duration) from manifest.json and run_results.json into each model’s side panel. For more information, see Monitor dbt Projects on Snowflake and Workspaces for dbt Projects on Snowflake.

View column-level lineage¶

Each model node in the DAG supports column-level lineage powered by Snowflake Horizon Catalog. Select Show columns on any model node to expand its column list. Selecting a column highlights all upstream and downstream models in the DAG that use that column, so you can trace how data flows through your pipeline.

Note

The project must be run at least once before column-level lineage is available. For more information, see Limitations for column-level lineage.

Execute models from the DAG¶

You can execute a subset of your dbt project object directly from the DAG by selecting the … menu on a model node. The following execution options are available:

Menu option	What it executes	Equivalent `--select` flag
Execute model	Only the selected model	`--select model_name`
Execute model+	The model and all downstream dependents	`--select model_name+`
Execute +model	The model and all upstream parents	`--select +model_name`
Execute +model+	The model, its parents, and its children	`--select +model_name+`

Selecting any option opens the Execute dbt project dialog with the Additional flags field pre-filled with the corresponding --select value. From the dialog, you can:

Choose the operation, such as Run, Test, or Build.
Choose the profile target (for example, dev or prod).
Edit the flags before executing if you want to refine the selection.

You can use the same --select syntax with the + graph operators in SQL and the Snowflake CLI:

EXECUTE DBT PROJECT my_dbt_project
  ARGS = 'build --select +stg_customers+ --target dev';

For more information about supported dbt commands and flags, see Supported dbt commands and flags.

View dbt project object properties¶

View the metadata Snowflake stores about a dbt project object to see what it’s called, who owns it, which version is the default, and where its files live in Snowflake’s internal snow://dbt/... stage.

To view the properties (such as name, owner, comment) of a specific dbt project object, use the DESCRIBE DBT PROJECT command, as shown in the following example:

DESCRIBE DBT PROJECT my_dbt_project;

The output shows the object’s name, owner, comment, versioning details, and external access integration. For more information, see DESCRIBE DBT PROJECT.

View all dbt projects¶

Use SHOW DBT PROJECTS when you want to see all dbt project objects you can access, plus key metadata.

SHOW DBT PROJECTS IN DATABASE mydb;

The output shows each object’s database, schema, owner, comment, when it was created and last updated, versioning details, and external access integration. For more information, see SHOW DBT PROJECTS.

Alternatively, use the snow dbt list command. For more information, see Listing all available dbt project objects.