Cognite Data Fusion CLI

Beta

The features described in this section are currently in beta testing and are subject to change. We recommend that you don't use the features in production systems.

Use the Cognite Data Fusion (CDF) CLI to manage data models from the command line.

Install the CDF CLI

Prerequisites

To install the CDF CLI, you must first install Node v16 or higher. For more information, see Node JS.

Follow the steps below to install the CDF CLI.

Install the CDF CLI globally

1. To install the CDF CLI globally, run the following command from a terminal:

   npm

   npm install --global @cognite/cdf-cli

   Yarn

   yarn global add @cognite/cdf-cli

Install the CDF CLI locally (JavaScript)

1. To install the CLI locally, run the following command from a terminal:

npm

npm install @cognite/cdf-cli

Yarn

yarn add @cognite/cdf-cli

pnpm

pnpm add @cognite/cdf-cli

Installing the CLI globally

You can also install the CLI globally and skip the project setup below with npm install -g @cognite/cdf-cli or yarn global add @cognite/cdf-cli

Use the CDF CLI easily in a javascript project

If you need to use the CLI in your project, add it within the scripts section in your package.json file.

{
  "name": "xxx",
  ...
  "scripts":{
      ...
      "cdf": "cdf"
      ...
  }
}

This will allow you to run npm run cdf or yarn cdf within your project.

CDF CLI commands

Usage: cdf [command] [subcommand] [option]

To view available commands, enter cdf --help.

The tables below describe the available commands and options.

Command	Alias	Description
signin [project]	login	Sign in to Cognite Data Fusion.
data-models [command]	dm	Manage data models to store and retrieve data.
signout	logout	Sign out current user.
status		Check if current user is signed in.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--version	[boolean]	Show the version number.
--help	[boolean]	Show help.

signin

Use the signin command to sign in to a CDF project.

Usage: cdf signin [project]

Command	Description
signin [project], [p]	Specifies the CDF project to sign in to (supports Microsoft Entra ID (formerly Azure Active Directory) currently) through interactive login or a client ID and client secret for machines.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.
--client-id	[string]	The client ID for the Microsoft Entra ID application.
--client-secret	[string]	The client secret for the Microsoft Entra ID application generated by the user. Secrets have a long expiry and are suitable for CI/CD.
--tenant	[string]	The Microsoft Entra ID tenant URI or ID.
--cluster	[string][default:api]	The CDF cluster name.
--devicecode	[boolean][default:false]	For interactive sign-in, shows a QR code to sign in. This is the recommended method for machines without a web browser.

data-models

Use that data-models command to create, publish, or delete a data model, retrieve data, and generate GraphQL code.

Usage: cdf data-models <command>

Command	Alias	Description
data-models create	dm create	Create a data model.
data-models publish	dm publish	Publish a data model.
data-models generate	dm generate	Generate a GraphQL client code for the schema you provide by fetching the introspection query from the server.
data-models delete	dm delete	Delete a data model.
data-models list	dm list	List all data models.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--version	[boolean]	Show the version number.
--help	[boolean]	Show help.

data-models create

Use the data-models create command to create a data model in a CDF project.

Usage: cdf data-models create <name>

Command	Type, default value	Description
data-models create [name]	[string][required]	The name of the data model.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.
--external-id	[string][default:"name"]	The external id of the data model.
--space	[string][default:"external-id"]	The space that the data model belongs in. Will create the space if it does not already exist

data-models publish

Use the data-models publish command to update a data model with a new GraphQL definition. If there are any breaking changes, a new version of the data model is published. For non-breaking changes, the existing version of the data model is updated.

Usage: cdf data-models publish

Command	Description
data-models publish	Updates a data model with a new GraphQL definition.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.
--external-id	[string]	The external id of the data model.
--file	[string]	The file containing the new GraphQL definition.
--space	[string][default:"external-id"]	The space that the data model belongs in.
--version	[string]	Data model version in.
--dry-run	[boolean]	Perform a dry run. Will only validate the data model without publishing.

data-models delete

Use the data-models delete command to delete the data model in CDF. The command will delete all unreferenced and unused data type within the data model.

Usage: cdf data-models delete

warning

This is an irreversible action!

Command	Description
data-models delete	Deletes a data model.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.
--external-id	[string]	The external id of the data model.
--space	[string][default:"external-id"]	The space that the data model belongs in.

data-models list

Use the data-models list command to list all data models in a CDF project.

In the terminal of your choice, enter the command below:

Usage: cdf data-models list

Command	Description
data-models list	Lists all data models.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.

data-models generate

Use the data-models generate command to generate a GraphQL client code for the provided schema by fetching the introspection query from the server.

Usage: cdf data-models generate

Command	Description
data-models delete	Generates a GraphQL client code.

Option	Type, default value	Description
-v, --verbose	[boolean][default:false]	Show debug messages.
--interactive	[boolean][default:true]	Request for interactive input if any required field is missing from options.
--help	[boolean]	Show help.
--external-id	[string]	The external id of the data model.
--plugins	[array]["typescript", "typescript-operations", "typescript-resolvers", "typescript-react-apollo", "typescript-apollo-angular"]	The code generation plugin. Separate multiple plugins with a space.
--operations-files	[string][default:"src/operations.graphql"]	The file path to scan for GraphQL code.
--output-file, -f	[string][default:"generated-types.ts"]	The file name of the generated code.
--space	[string][default:"external-id"]	The space that the data model belongs in.
--version	[string]	Data model version in.

signout

Use the signout command [alias:logout] to sign out the current user.

Usage: cdf signout

status

Use the status command to check if the current user is signed in.

Usage: cdf status

The output fails if the user isn't signed in.