preset

Presets let teams encode project standards into a shareable YAML file. Use them with dbt-forge init --preset to enforce adapter choices, package selections, and feature flags across multiple dbt projects.

Commands

# Validate a preset file
dbt-forge preset validate <path-or-url>

# Use a preset during init
dbt-forge init my_project --preset company-standard.yml

Preset format

A preset is a YAML file with three sections:

name: "Company Standard"
description: "Standard config for the analytics team"

defaults:
  adapter: "Snowflake"
  marts: ["finance", "marketing", "product"]
  packages: ["dbt-utils", "dbt-expectations", "elementary"]
  add_examples: true
  add_sqlfluff: true
  ci_providers: ["GitHub Actions"]
  add_unit_tests: false
  add_metricflow: false
  add_snapshot: true
  add_seed: false
  add_exposure: true
  add_macro: false
  add_pre_commit: true
  add_env_config: true
  team_owner: "@analytics-team"

locked:
  - adapter
  - add_sqlfluff
  - ci_providers

name and description

Optional metadata for the preset. Not used during scaffolding but helps identify the preset when sharing across teams.

defaults

Values that override the default prompt selections during init. The user still sees the prompt and can change the value, unless the field is also listed in locked.

locked

Fields listed here skip the interactive prompt entirely and use the value from defaults. Use this to enforce standards that individual users should not change.

Every locked field must have a corresponding entry in defaults. A locked field without a default is a validation error.

Valid fields

These fields can appear in defaults and locked:

Field	Type	Description
adapter	string	Warehouse adapter. One of: BigQuery, Snowflake, PostgreSQL, DuckDB, Databricks, Redshift, Trino, Spark
marts	list of strings	Mart domains to scaffold (e.g., ["finance", "marketing"])
packages	list of strings	Starter packages from the curated registry
add_examples	boolean	Generate example staging models and tests
add_sqlfluff	boolean	Generate .sqlfluff and .sqlfluffignore
ci_providers	list of strings	CI providers. Values: "GitHub Actions", "GitLab CI", "Bitbucket Pipelines"
add_unit_tests	boolean	Generate unit test examples (requires add_examples)
add_metricflow	boolean	Generate MetricFlow semantic model example
add_snapshot	boolean	Generate example snapshot
add_seed	boolean	Generate example seed CSV and YAML
add_exposure	boolean	Generate example exposure YAML
add_macro	boolean	Generate example macro
add_pre_commit	boolean	Generate pre-commit hooks and .editorconfig
add_env_config	boolean	Generate .env.example and generate_schema_name.sql
team_owner	string	GitHub team or user for CODEOWNERS (e.g., "@analytics-team")

Unknown fields in defaults or locked cause a validation error.

preset validate

dbt-forge preset validate company-standard.yml
dbt-forge preset validate https://example.com/presets/standard.yml

Checks a preset file for:

• Unknown fields in defaults or locked
• Locked fields that have no corresponding default value
• Invalid adapter names (must be one of the 8 supported adapters)
• Invalid CI provider names (must be one of the 3 supported providers)

Exits with code 0 if the preset is valid, code 1 if there are errors.

Example output for an invalid preset:

Preset validation errors:
  - Unknown field in defaults: 'warehouse'
  - Locked field 'adapter' has no value in defaults
  - Invalid adapter: 'MySQL'

Using presets with init

# Local file
dbt-forge init analytics_core --preset company-standard.yml

# HTTPS URL
dbt-forge init analytics_core --preset https://example.com/presets/standard.yml

# Combined with --defaults (no prompts shown)
dbt-forge init analytics_core --preset company-standard.yml --defaults

The --preset flag accepts a local file path or an HTTPS URL. The preset is validated before init begins. If validation fails, init exits without scaffolding.

Interactive mode (default)

When used without --defaults:

• Locked fields are skipped entirely — the preset value is used
• Default fields pre-populate the prompt — the user can change them
• Unset fields use the normal dbt-forge defaults

Non-interactive mode (--defaults)

When combined with --defaults:

• Preset values override the built-in defaults
• No prompts are shown
• Locked fields behave the same as default fields (both are applied without prompting)

Behavior and limits

• Presets only affect the init command. They do not change add, doctor, or status behavior.
• HTTPS URLs are fetched with Python’s urllib. No authentication is supported.
• The preset file must be valid YAML. Malformed files cause an error before init starts.
• Locked fields must have a value in defaults. A locked field without a default is a validation error.
• The project_name field cannot be set via presets — it is always provided as a CLI argument or prompted.