A batteries-included GitHub Codespace for dbt development — no local install required.
Opens a VS Code environment with the dbt Fusion engine (Rust-based) and the official dbt VS Code extension pre-installed.
Start a new dbt project — creates a repo in your org from this template, then open a Codespace from your new repo via Code → Codespaces:
Try it out — opens a Codespace directly from this template without creating a repo:
Codespaces work best in Chrome or Firefox. Safari users may hit Service Worker errors — enable third-party cookies if so.
Before creating your Codespace, add these repository secrets so credentials are injected automatically at startup:
| Secret | Description |
|---|---|
SNOWFLAKE_ACCOUNT |
Account identifier, e.g. abc12345.us-east-1 |
SNOWFLAKE_USER |
Your Snowflake username |
SNOWFLAKE_PASSWORD |
Your Snowflake password |
SNOWFLAKE_DATABASE |
Target database |
SNOWFLAKE_WAREHOUSE |
Virtual warehouse name |
SNOWFLAKE_SCHEMA |
Dev schema (defaults to dev) |
SNOWFLAKE_ROLE |
Role (optional) |
Set them at: Repository → Settings → Secrets and variables → Codespaces
Docs: Specifying recommended secrets for a repository
The profiles.yml at the root already references the secrets above. Update dbt_project.yml with your project name and adjust the model configurations to match your layering convention.
dbt debug # verify connection
dbt deps # install packages
dbt run # execute models
dbt test # run tests| Component | Detail |
|---|---|
| Engine | dbt Fusion (Rust binary — fast, no Python dependency) |
| Extension | dbt Labs VS Code extension with DAG explorer, model preview, autocomplete |
| profiles.yml | Snowflake template wired to Codespaces secrets via env_var() |
| dbt_project.yml | Minimal skeleton with staging (views) and marts (tables) layers |
Swap out the outputs.dev block in profiles.yml for any adapter dbt Fusion supports. See dbt connection profiles docs.