Skip to content
dbt-forge

dbt project template

Teams often search for a dbt project template when they want a repeatable starting point instead of rebuilding the same folders, profiles, and setup files by hand. dbt-forge acts as a dbt project scaffold CLI: it generates the core structure, adapter-aware configuration, starter models, and optional tooling that most teams add to every new project anyway.

What a good dbt project template should include

Section titled “What a good dbt project template should include”

A practical dbt project template should give you enough structure to start shipping models immediately without locking you into a rigid architecture.

At minimum, the template should cover:

  • dbt project configuration with dbt_project.yml
  • dependency management for the chosen adapter
  • a consistent models/ layout for staging, intermediate, and marts
  • warehouse connection profiles that stay out of version control
  • package definitions with pinned versions
  • starter tests, selectors, and documentation files
  • optional CI, linting, and pre-commit setup for team workflows

dbt-forge scaffolds that baseline with init and then extends the same project later with add.

Why teams use a dbt scaffold instead of copying old repos

Section titled “Why teams use a dbt scaffold instead of copying old repos”

Copying an old repository usually drags along warehouse-specific settings, stale packages, inconsistent naming, and files that nobody still understands. A dedicated dbt scaffold is safer because it starts from templates with explicit defaults.

dbt-forge improves on copy-paste templates by:

  • rendering adapter-aware profiles
  • tracking generated files for later updates with update
  • validating the finished project with doctor
  • supporting generated sources from live warehouse metadata via add source --from-database

Typical dbt-forge project template output

Section titled “Typical dbt-forge project template output”

After running dbt-forge init my_project --defaults, the generated template usually includes:

  • dbt_project.yml
  • pyproject.toml
  • profiles/profiles.yml
  • packages.yml
  • selectors.yml
  • models/staging/, models/intermediate/, and models/marts/
  • .dbt-forge.yml for template update tracking
  • optional SQLFluff, CI, pre-commit, and environment config files

Use Project structure for a full breakdown of each file and directory.

When to customize the template

Section titled “When to customize the template”

The scaffold should be the starting point, not the final architecture. Most teams customize it after generation to:

  • add domain-specific marts
  • change package choices
  • enable Mesh-specific project boundaries
  • add environment-specific macros and CI rules
  • remove example content before production use

That is why the recommended flow is:

  1. Follow Getting started to install the CLI and generate the initial project.
  2. Use init to create the first scaffold.
  3. Use add to extend the project with marts, models, tests, and packages.
  4. Run doctor to validate naming, test coverage, and health checks.

Next steps

Section titled “Next steps”