Find model impact

When to use this

Use this recipe before changing a dbt model to understand which downstream models and tests depend on it. This helps you assess the blast radius of a change before it reaches CI or production.

Inputs required

• manifest.json (run results not required for lineage analysis)
• Optional: run_results.json (adds execution status to downstream nodes)

Recommended interface

Interface	Use when
CLI	You need dependency lists as JSON for scripting or pre-change review
Web	You want to visually explore the lineage graph around a model
MCP	An AI agent needs to reason about impact scope and suggest test coverage

Step 1: Find the model by name

If you only have a partial name:

npx @dbt-tools/cli discover --dbt-target ./target "orders" --limit 5 --json

Note the unique_id from the output. It will look like model.my_project.orders.

Step 2: Inspect downstream dependencies

npx @dbt-tools/cli deps model.my_project.orders --dbt-target ./target --direction downstream --json

The output lists all models, tests, snapshots, and exposures that depend—directly or transitively—on the target model.

Step 3: Inspect upstream dependencies

npx @dbt-tools/cli deps model.my_project.orders --dbt-target ./target --direction upstream --json

Upstream dependencies show what the model reads. If you are changing a source, this helps you find all models that consume it.

Step 4: Explain the model

npx @dbt-tools/cli explain model.my_project.orders --dbt-target ./target --json

The explain output includes column definitions, tests associated with the model, and configuration metadata. Check whether there are downstream exposures or BI tools connected to this model.

Step 5 (optional): Open in Web for visual lineage

npx @dbt-tools/web --dbt-target ./target

The Web UI shows the lineage graph visually. Search for the model by name and click to expand its upstream and downstream nodes. This is especially useful for models with many indirect dependents.

Reading the output

Field in deps output	What it means
Direct downstream nodes	Models or tests that directly reference this model
Transitive downstream nodes	Models that depend on direct dependents
Exposures	BI reports, dashboards, or other downstream consumers declared in exposure blocks
Tests	Tests that run against this model or its columns

Common questions

"What breaks if I rename this model?" Run deps --direction downstream and count the transitive dependents. Any model in the list will break if the source is renamed without also updating the ref() call.

"Is this model used by anyone?" If deps --direction downstream returns an empty list, the model has no declared dependents in the project. It may still be referenced by other tools or by external consumers not captured in the manifest.

"What sources does this model read?" Run deps --direction upstream. Source nodes appear with unique_id values like source.my_project.raw.orders.

Related

• Debug a failed run — combine impact analysis with failure context
• Open CLI result in Web — move from JSON to browser for visual lineage
• CLI cheatsheet
• Deep links — link from CLI JSON to specific Web views