Skip to main content

Configure and use the dbt Cloud CLI

Learn how to configure the dbt Cloud CLI for your dbt Cloud project to run dbt commands, like dbt environment show to view your dbt Cloud configuration or dbt compile to compile your project and validate models and tests. You'll also benefit from:

  • Secure credential storage in the dbt Cloud platform.
  • Automatic deferral of build artifacts to your Cloud project's production environment.
  • Speedier, lower-cost builds.
  • Support for dbt Mesh (cross-project ref), and more.

Prerequisites

  • You must set up a project in dbt Cloud.
    • Note — If you're using the dbt Cloud CLI, you can connect to your data platform directly in the dbt Cloud interface and don't need a profiles.yml file.
  • You must have your personal development credentials set for that project. The dbt Cloud CLI will use these credentials, stored securely in dbt Cloud, to communicate with your data platform.
  • You must be on dbt version 1.5 or higher. Refer to dbt Cloud versions to upgrade.

Configure the dbt Cloud CLI

Once you install the dbt Cloud CLI, you need to configure it to connect to a dbt Cloud project.

  1. In dbt Cloud, navigate to Develop and click Configure dbt Cloud CLI to download your dbt_cloud.yml credentials file.

    Details

    Region URLs to download credentials You can also download the credentials from the links provided based on your region:

  2. Save the dbt_cloud.yml file in the .dbt directory, which stores your dbt Cloud CLI configuration. Store it in a safe place as it contains API keys. Check out the FAQs to learn how to create a .dbt directory and move the dbt_cloud.yml file.

  3. Follow the banner instructions and download the config file to:

    • Mac or Linux: ~/.dbt/dbt_cloud.yml
    • Windows: C:\Users\yourusername\.dbt\dbt_cloud.yml

The config file looks like this:

version: "1"
context:
active-project: "<project id from the list below>"
active-host: "<active host from the list>"
defer-env-id: "<optional defer environment id>"
projects:
- project-id: "<project-id>"
account-host: "<account-host>"
api-key: "<user-api-key>"

- project-id: "<project-id>"
account-host: "<account-host>"
api-key: "<user-api-key>"
  1. After downloading the config file and creating your directory, navigate to a dbt project in your terminal:

    cd ~/dbt-projects/jaffle_shop
  2. In your dbt_project.yml file, ensure you have or include a dbt-cloud section with a project-id field. The project-id field contains the dbt Cloud project ID you want to use.

    # dbt_project.yml
    name:
    version:
    # Your project configs...

    dbt-cloud:
    project-id: PROJECT_ID
    • To find your project ID, select Develop in the dbt Cloud navigation menu. You can use the URL to find the project ID. For example, in https://YOUR_ACCESS_URL/develop/26228/projects/123456, the project ID is 123456.
  3. You should now be able to use the dbt Cloud CLI and run dbt commands like dbt environment show to view your dbt Cloud configuration details or dbt compile to compile models in your dbt project.

With your repo recloned, you can add, edit, and sync files with your repo.

Set environment variables

To set environment variables in the dbt Cloud CLI for your dbt project:

  1. Select the gear icon on the upper right of the page.
  2. Then select Profile Settings, then Credentials.
  3. Click on your project and scroll to the Environment Variables section.
  4. Click Edit on the lower right and then set the user-level environment variables.

Use the dbt Cloud CLI

The dbt Cloud CLI uses the same set of dbt commands and MetricFlow commands as dbt Core to execute the commands you provide. For example, use the dbt environment command to view your dbt Cloud configuration details. With the dbt Cloud CLI, you can:

  • Run multiple invocations in parallel and ensure safe parallelism, which is currently not guaranteed by dbt-core.
  • Automatically defers build artifacts to your Cloud project's production environment.
  • Supports project dependencies, which allows you to depend on another project using the metadata service in dbt Cloud.
    • Project dependencies instantly connect to and reference (or ref) public models defined in other projects. You don't need to execute or analyze these upstream models yourself. Instead, you treat them as an API that returns a dataset.
Use the --help flag

As a tip, most command-line tools have a --help flag to show available commands and arguments. Use the --help flag with dbt in two ways:

  • dbt --help: Lists the commands available for dbt
  • dbt run --help: Lists the flags available for the run command

Lint SQL files

From the dbt Cloud CLI, you can invoke SQLFluff which is a modular and configurable SQL linter that warns you of complex functions, syntax, formatting, and compilation errors. Many of the same flags that you can pass to SQLFluff are available from the dbt Cloud CLI.

The available SQLFluff commands are:

  • lint — Lint SQL files by passing a list of files or from standard input (stdin).
  • fix — Fix SQL files.
  • format — Autoformat SQL files.

To lint SQL files, run the command as follows:

dbt sqlfluff lint [PATHS]... [flags]

When no path is set, dbt lints all SQL files in the current project. To lint a specific SQL file or a directory, set PATHS to the path of the SQL file(s) or directory of files. To lint multiple files or directories, pass multiple PATHS flags.

To show detailed information on all the dbt supported commands and flags, run the dbt sqlfluff -h command.

Considerations

When running dbt sqlfluff from the dbt Cloud CLI, the following are important behaviors to consider:

  • dbt reads the .sqlfluff file, if it exists, for any custom configurations you might have.
  • For continuous integration/continuous development (CI/CD) workflows, your project must have a dbt_cloud.yml file and you have successfully run commands from within this dbt project.
  • An SQLFluff command will return an exit code of 0 if it ran with any file violations. This dbt behavior differs from SQLFluff behavior, where a linting violation returns a non-zero exit code. dbt Labs plans on addressing this in a later release.

FAQs

 How to create a .dbt directory and move your file
 How to skip artifacts from being downloaded
0