dbt-score

A comprehensive linter for dbt metadata that helps maintain high-quality data models at scale.

dbt-score lint
🥉 orders (score: 2.7)
  WARN (medium) dbt_score.rules.generic.columns_have_description: Columns lack a description: customer_id, customer_name.
  WARN (high) dbt_score.rules.generic.has_description: Model lacks a description.
  WARN (medium) dbt_score.rules.generic.has_owner: Model lacks an owner.
  WARN (medium) dbt_score.rules.generic.sql_has_reasonable_number_of_lines: SQL query too long: 238 lines (> 200).
  WARN (medium) dbt_score_rules.custom_rules.has_test: Model lacks a test.

What is dbt-score?

dbt-score is a powerful linting tool designed to evaluate and score dbt (Data Build Tool) models based on metadata quality. It helps data teams maintain consistent standards across dbt projects by programmatically enforcing best practices for documentation, testing, naming conventions, and more.

Key Features

• 🔍 Comprehensive Linting: Evaluates dbt entities against configurable rules for documentation, tests, naming, and structure
• 📊 Scoring System: Provides numerical scores (0-10) for individual models and overall project health
• 🎯 Flexible Configuration: Customizable rules, severity levels, and scoring thresholds via pyproject.toml
• 🚀 CI/CD Integration: Fail builds when quality standards aren't met
• 📈 Progress Tracking: Visual badges and scoring to track data quality improvements over time
• 🔧 Extensible: Create custom rules tailored to organization-specific needs

Quick Start

Installation

pip install dbt-score

Note: Install dbt-score in the same environment as dbt-core.

Basic Usage

Run dbt-score from your dbt project root:

# Basic linting
dbt-score lint

# Also show passing tests
dbt-score lint --show all

# Lint specific models
dbt-score lint --select +my_model+

# Auto-generate manifest (via `dbt parse`) and lint
dbt-score lint --run-dbt-parse

Example Output

dbt-score lint --show all
🥉 orders (score: 2.7)
  WARN (medium) dbt_score.rules.generic.columns_have_description: Columns lack a description: customer_id, customer_name.
  WARN (high) dbt_score.rules.generic.has_description: Model lacks a description.
  WARN (medium) dbt_score.rules.generic.has_owner: Model lacks an owner.
  WARN (medium) dbt_score.rules.generic.sql_has_reasonable_number_of_lines: SQL query too long: 238 lines (> 200).
  WARN (medium) dbt_score_rules.custom_rules.has_test: Model lacks a test.

🥇 customers (score: 10.0)
  OK    dbt_score.rules.generic.columns_have_description
  OK    dbt_score.rules.generic.has_description
  OK    dbt_score.rules.generic.has_owner
  OK    dbt_score.rules.generic.sql_has_reasonable_number_of_lines
  OK    dbt_score_rules.custom_rules.has_test

Project score: 6.3 🥈

Configuration

Configure dbt-score via pyproject.toml in the dbt project root:

[tool.dbt-score]
# Fail CI if project score falls below threshold
fail_project_under = 7.5
fail_any_item_under = 8.0

# Disable specific rules
disabled_rules = ["dbt_score.rules.generic.columns_have_description"]

# Configure badges
[tool.dbt-score.badges]
first.threshold = 10.0
first.icon = "🥇"
second.threshold = 8.0
second.icon = "🥈"
third.threshold = 6.0
third.icon = "🥉"
wip.icon = "🏗️"

# Customize rule severity and parameters
[tool.dbt-score.rules."dbt_score.rules.generic.sql_has_reasonable_number_of_lines"]
severity = 1
max_lines = 300

Why Use dbt-score?

As dbt projects grow to hundreds or thousands of models, maintaining consistent metadata becomes increasingly challenging:

• Inconsistent Documentation: Some models are well-documented, others lack basic descriptions
• Missing Tests: Critical models without proper data quality tests
• Naming Inconsistencies: Models that don't follow established conventions
• Technical Debt: Long, complex SQL queries that are hard to maintain
• Compliance Issues: Missing ownership or governance metadata

dbt-score addresses these challenges by:

• Automated Quality Checks: Continuously evaluate dbt projects against best practices
• Objective Scoring: Get clear, numerical feedback on model quality
• Team Alignment: Establish shared standards across data teams
• CI/CD Integration: Prevent quality regressions in production

Built-in Rules

dbt-score comes with a small set of rules covering needs applicable to most dbt projects.

Advanced Usage

Custom Rules

Create organization-specific rules by writing simple Python functions:

from dbt_score import Model, rule, RuleViolation

@rule
def model_has_business_owner(model: Model) -> RuleViolation:
    if model.meta.get("business_owner") is None:
        return RuleViolation("Model lacks a business owner.")

CI/CD Integration

Add dbt-score to CI pipelines:

- name: Run dbt-score
  run: |
    dbt-score lint --run-dbt-parse

or equivalent in your favourite CI platform. dbt-score exits with 0 or 1 to signal success or failure, making integrations a breeze!

Selective Linting

Use dbt's selection syntax to lint specific parts of projects:

# Lint only staging models
dbt-score lint --select staging.*

# Lint a model and its dependencies
dbt-score lint --select +my_important_model

# Lint recently changed models
dbt-score lint --select state:modified

Documentation

For comprehensive documentation, including detailed rule descriptions, configuration options, and advanced usage patterns, visit the dbt-score documentation website.

Contributing

Contributions are welcome! This includes:

• Reporting bugs or requesting features
• Improving documentation
• Adding new rules or formatters
• Fixing issues

Check out the contributing guide to get started. 🚀

Requirements

• Python 3.10+
• dbt-core 1.5+

License

This project is licensed under the MIT License - see the LICENSE.txt file for details.

---