docglow 0.1.0

Next-generation dbt documentation site generator

docglow

Next-generation documentation site generator for dbt Core projects.

Docglow replaces the default dbt docs generate + dbt docs serve workflow with a modern, interactive single-page application featuring lineage exploration, column-level docs, test results, and project health scoring — at a fraction of the cost of enterprise data catalogs.

Install

pip install docglow

Quick Start

# Generate the site from your dbt project
docglow generate --project-dir /path/to/dbt/project --output-dir ./site

# Serve locally
docglow serve --dir ./site

Features

• Interactive lineage explorer — drag, filter, and explore upstream/downstream dependencies with configurable depth and layer visualization
• Column-level documentation — searchable column tables with descriptions, types, and test status
• Test results dashboard — per-model test outcomes with pass/fail/warn indicators
• Project health score — coverage metrics for descriptions, tests, and documentation completeness
• Full-text search — instant search across all models, sources, and columns
• Single static site — no backend required, deploy anywhere (S3, GitHub Pages, Netlify, etc.)

CLI Commands

Command	Description
docglow generate	Generate the documentation site from dbt artifacts
docglow serve	Serve the generated site locally
docglow health	Show project health score and coverage metrics
docglow profile	Run column-level profiling (requires docglow[profiling])
docglow publish	Publish to docglow.dev (requires docglow[cloud])

Configuration

Add a docglow.yml to your dbt project root for optional customization (layer definitions, display settings, etc.). Docglow works out of the box without any configuration — just point it at a dbt project with compiled artifacts in target/.

Requirements

• Python 3.10+
• A dbt project with target/manifest.json (run dbt compile or dbt run first)

License

MIT