dbt.build high conf mediumOBO
Executes models, tests, snapshots, and seeds across the whole DAG in dependency order, rewriting production tables project-wide in a single invocation.
dbt.trigger_job_run high conf mediumhuman approval
Triggers a production job run and can override the git branch, target schema, or other settings, letting one principal deploy and execute pipeline code outside normal change control.
All other tools (19)
dbt.cancel_job_run medium conf mediumaudit
Cancels an in-flight job run, interrupting a scheduled production pipeline and potentially leaving downstream tables stale.
dbt.clone medium conf mediumaudit
Clones selected nodes from a source state into target schemas, duplicating tables and any personal data they hold into a new location.
dbt.compile low conf mediumallow
Renders executable SQL from models, tests, and analyses to validate Jinja logic without running it against the database.
dbt.docs low conf mediumallow
Generates the project documentation site from model and column metadata; no impact on warehouse data.
dbt.generate_model_yaml low conf mediumallow
Generates model YAML with columns, optionally inheriting upstream descriptions; a scaffolding write to project files, not to data.
dbt.generate_source low conf mediumallow
Introspects database schemas to generate source YAML, reading table and column names from the warehouse into project files.
dbt.generate_staging_model low conf mediumallow
Generates staging model SQL from a source table, producing boilerplate transformation code for review before it is run.
dbt.get_dimension_values medium conf mediumaudit
Returns the distinct values of a dimension column, directly exposing raw field contents such as customer names, emails, or region codes.
dbt.get_job_details low conf mediumallow
Returns a job's configuration including its triggers, schedule, and dbt commands, exposing how the production pipeline is orchestrated.
dbt.get_job_run_error low conf mediumallow
Retrieves error and warning details for a job run, which may echo query fragments or data values from the failing step.
dbt.get_model_details low conf mediumallow
Returns a model's compiled SQL, columns, and schema, revealing the structure of the underlying data warehouse tables.
dbt.list_job_run_artifacts low conf mediumallow
Lists the artifacts a job run produced, such as run results and compiled catalogs that describe the project's data structures.
dbt.parse low conf mediumallow
Parses and validates project files for syntax correctness without touching the database.
dbt.query_metrics medium conf mediumaudit
Executes Semantic Layer metric queries with filtering and grouping, returning aggregated business data that can be sliced by dimensions carrying personal identifiers.
dbt.retry_job_run medium conf mediumaudit
Re-executes a failed job run, re-materializing production models without a fresh review of the failure cause.
dbt.run medium conf mediumaudit
Materializes selected models in the database, overwriting the target tables with freshly transformed data.
dbt.show medium conf mediumaudit
Executes arbitrary SQL against the database and returns a result preview, surfacing raw column values that may include personal data.
dbt.test low conf mediumallow
Runs data and model integrity tests against the warehouse; read-only validation that materializes no new state.
dbt.text_to_sql low conf mediumallow
Generates SQL from a natural-language prompt using project context without executing it; output must still be reviewed before it touches data.