What is the difference between dbt unit tests and data tests?

Unit tests validate your SQL logic using static, mocked inputs before the model is built, whereas data tests validate the actual data produced in your warehouse after the model is materialized.

Can I use this skill for ephemeral models?

Yes, the skill provides specific guidance on using the SQL format for mocking inputs, which is a requirement when unit testing models that depend on ephemeral models.

Does it support mocking data in different formats?

Yes, it supports mocking data using YAML dictionaries (default), CSV strings/files, and SQL queries/files, depending on the complexity of your data types.

Can I test incremental models with this skill?

Yes, it includes patterns for testing what logic will be merged or inserted during an incremental run, though it does not test the final state of the database table itself.

dbt Unit Test Generator

Name: dbt Unit Test Generator
Author: dbt-labs

bydbt-labs

•

304

•

Database Management

Generates standardized dbt unit test YAML definitions to validate SQL transformation logic using mocked inputs and expected outputs.

This skill streamlines the implementation of test-driven development (TDD) within dbt projects by helping developers create robust unit tests. It automates the generation of the Model-Input-Output (MIO) pattern, allowing users to mock upstream dependencies like sources, seeds, and models without requiring actual warehouse data. By facilitating the validation of complex SQL logic—such as window functions, regex, and intricate joins—before materialization, it ensures high data quality and prevents regressions in production environments.

Key Features

01Mocks upstream dependencies including refs, sources, and seeds.

02Supports multiple data formats including YAML dict, CSV, and SQL queries.

03Generates comprehensive unit test YAML structures for dbt models.

04Provides warehouse-specific guidance for BigQuery, Snowflake, Postgres, and Spark.

05Handles specialized scenarios like incremental models and ephemeral dependencies.

06304 GitHub stars

Use Cases

01Validating complex SQL transformations involving window functions or date math.

02Creating regression tests to ensure bug fixes remain effective over time.

03Implementing test-driven development (TDD) for new dbt models.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dbt-labs/dbt-agent-skills adding-dbt-unit-test

For use in Claude.ai and ChatGPT

Download Skill