Data Validation Tool (dvt)

CLI tool for dbt data validation workflows. Born from the scripts folder of dbt-audit-helper-ext, this tool consolidates validation utilities into a proper Python package.

Table of Contents

• Data Validation Tool (dvt)
  • Table of Contents
  • Installation
  • Quick Start
  • Commands
    • dvt init
    • dvt run
    • dvt cloud
  • Development
  • License

Installation

pip install data-validation-tool

Or with uv:

uv pip install data-validation-tool

Quick Start

# Generate validation macros for your dbt models
dvt init --models-dir models/03_mart

# Run validations
dvt run --models-dir models/03_mart --type all

# Generate dbt Cloud job configurations
export DBT_CLOUD_ACCOUNT_ID=12345
export DBT_CLOUD_PROJECT_ID=67890
export DBT_CLOUD_ENVIRONMENT_ID=11111
dvt cloud --models-dir models/03_mart

Commands

dvt init

Generate validation macros for dbt models. Scans your models directory and creates validation macro files compatible with dbt-audit-helper-ext.

dvt init [OPTIONS]

Options:
  -d, --models-dir PATH   Directory containing dbt models (default: models/03_mart)
  -m, --model TEXT        Specific model name to generate macros for
  -o, --output-dir PATH   Output directory for generated macros (default: macros/validation)
  --help                  Show this message and exit

dvt run

Execute validation processes. Runs dbt models and executes validation macros to compare source and target data.

dvt run [OPTIONS]

Options:
  -d, --models-dir PATH   Directory containing dbt models (default: models/03_mart)
  -m, --model TEXT        Specific model name to validate
  -t, --type TYPE         Validation type: all, count, schema, all_row, all_col (default: all)
  -p, --audit-date TEXT   Audit helper date for cloning from legacy data
  -r, --skip-run          Skip model runs, validate only
  -v, --run-only          Run models only, skip validation
  --help                  Show this message and exit

dvt cloud

Generate dbt Cloud job configurations as YAML. Creates files compatible with dbt-jobs-as-code.

dvt cloud [OPTIONS]

Options:
  -d, --models-dir PATH      Directory containing dbt models (default: models/03_mart)
  -m, --model TEXT           Specific model name to create job for
  -o, --output PATH          Output path for jobs YAML (default: dataops/dbt_cloud_jobs.yml)
  --account-id TEXT          dbt Cloud account ID (or set DBT_CLOUD_ACCOUNT_ID) [required]
  --project-id TEXT          dbt Cloud project ID (or set DBT_CLOUD_PROJECT_ID) [required]
  --environment-id TEXT      dbt Cloud environment ID (or set DBT_CLOUD_ENVIRONMENT_ID) [required]
  --help                     Show this message and exit

Development

Clone the repository and install development dependencies:

git clone https://github.com/infinitelambda/data-validation-tool.git
cd data-validation-tool
uv sync --all-extras

Run tests with coverage:

uv run pytest

Format and lint code:

uv run ruff format .
uv run ruff check .

License

Apache License 2.0 - see LICENSE for details.