dbt-colibri 0.3.6

A lightweight, developer-friendly CLI tool and self-hostable dashboard for extracting and visualizing column-level lineage from your dbt projects.

Built for data teams who want transparent, flexible lineage tracking without vendor lock-in or complex enterprise tooling.

Why dbt-colibri?

• 🔍 Complete visibility: Easy UI, track how every column flows through your dbt transformations
• ⚡ Fast & lightweight: Generate reports in seconds from your existing dbt artifacts
• 🏠 Self-hosted: No cloud dependencies or external services required

Documentation site: https://www.docs.colibri-data.com

Quick Start

Installation

# Using uv (recommended)
uv add dbt-colibri

# Using pip
pip install dbt-colibri

Basic Usage

1. Run dbt to generate the required artifacts:

   dbt compile
   dbt docs generate
   

2. Generate lineage report:

   colibri generate
   

3. View results: Open dist/index.html in your browser

That's it! Your column lineage dashboard is ready. Note you can also use dbt run/build, to generate the manifest.json.

Documentation

CLI Commands

colibri generate

Generates column lineage reports from your dbt project.

colibri generate [OPTIONS]

Options:

• --manifest: Path to dbt manifest.json (default: target/manifest.json)
• --catalog: Path to dbt catalog.json (default: target/catalog.json)
• --output-dir: Output directory (default: dist/)
• --help: Show help message
• --light: For very big dbt projects, excludes attributes like compiled SQL and returns smaller HTML file.

Output Files

• colibri-manifest.json: Lineage data
• index.html: Interactive (standalone) visualization dashboard

Project Structure

your-dbt-project/
├── target/
│   ├── manifest.json    # Generated by dbt
│   └── catalog.json     # Generated by dbt docs generate
└── dist/                # Generated by colibri
    ├── index.html       # Interactive dashboard
    └── colibri-manifest.json

Advanced Usage

CI/CD Integration

The easiest way to deploy your static html is through github/gitlab pages (if you are on enterprise license you can do this privately)

You can find the full example workflow at docs/github_pages_example.yml.

General idea

1. After every change to the production dbt code (push the main branch), GitHub Actions will:
   • Set up Python and install dependencies with uv.
   • Compile and generate docs needed for colibri.
   • Run colibri generate to build the static HTML report in the dist/ folder.
2. The dist/ folder is uploaded as an artifact and deployed natively to GitHub Pages using the official actions/deploy-pages action.
3. The result is available at your repository’s Pages URL.

Gitlab has similar functionality. Other options are writing the file to a bucket and mount it into a web server container (nginx).

Technical Details

Requirements

• Python: tested on versions 3.9, 3.11, 3.13

• Supported dbt Adapters:

  • Snowflake,
  • BigQuery,
  • Redshift,
  • duckDB,
  • Postgres
  • Databricks (limited to SQL models)
  • Athena
  • Trino
  • SQL Server (TSQL)
  • ClickHouse
  • Oracle
  • StarRocks

dbt Compatibility

Architecture

dbt-colibri leverages:

• SQLGlot for SQL parsing and column lineage extraction
• dbt artifacts (manifest.json, catalog.json) for metadata
• Static HTML/JS for zero-dependency dashboard deployment

Contributing

We welcome contributions! Raise an issue or request a feature, if you are open to contribute you can let us now in the issue.

• Issues: GitHub Issues
• Discussions: GitHub Discussions

Development Setup

# Clone the repository
git clone https://github.com/your-org/dbt-colibri.git
cd dbt-colibri

# Install development dependencies
uv sync --dev

# Run tests
pytest

# Format code
ruff format

License

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

Acknowledgments

This project builds upon excellent open source work:

• dbt-column-lineage-extractor - Original column lineage extraction logic
• SQLGlot - SQL parsing and transformation
• elementary-data - Inspiration for static HTML report structure