Skip to content
On this page

Command Line Interface

Directus has two command line interfaces (CLI) that you can use for various actions. One is used for server-side actions that relate to your on-prem instance, like migrating the database or resetting a user, while the other allows you to interact with a Directus instance as you would with an SDK.

Requirements

  • Node.js v18, specifically version 18.17 or higher.

Server

For server-side CLI, all functionality can be accessed by running npx directus <command> in your project folder.

Initialize a New Project

bash
npm init directus-project <project-folder>
npm init directus-project <project-folder>

Will setup a new Directus project based on the provided configuration values. The script installs the required database driver, creates a .env file and installs the required dependencies.

Bootstrap a Project

bash
npx directus bootstrap
npx directus bootstrap

Will use an existing .env file (or existing environment variables) to either install the database (if it's empty) or migrate it to the latest version (if it already exists and has missing migrations).

This is very useful to use in environments where you're doing standalone automatic deployments, like a multi-container Kubernetes configuration.

First User

You can use the ADMIN_EMAIL and ADMIN_PASSWORD environment variables to automatically provision the first user on first creation using the bootstrap command. See Environment Variables for more information.

Skip Admin User/Role

You can pass the --skipAdminInit option to bootstrap, if you're creating your Admin role/user in another way (with a custom migration or an external service, for example).

Install the Database

bash
npx directus database install
npx directus database install

Installs the initial Directus system tables on an empty database. Used internally by bootstrap.

It should be used only in specific cases, e.g. when you want to run something between install and migrate. You probably should call directus database migrate:latest afterwards manually.

You may want to use directus bootstrap instead.

Upgrade the Database

bash
npx directus database migrate:latest
npx directus database migrate:up
npx directus database migrate:down
npx directus database migrate:latest
npx directus database migrate:up
npx directus database migrate:down

Migrate the database up/down to match the versions of Directus. Once you update Directus itself, make sure to run npx directus database migrate:latest (or npx directus bootstrap) to update your database.

Migrate Schema to a different Environment

This allows you to do things like migrate a schema from development to production. To move your configured data model between Directus instances, you can use the schema "snapshot" and "apply" commands.

Snapshot the Data Model

Directus can automatically generate a snapshot of your current data model in YAML or JSON format. This includes all collections, fields, and relations, and their configuration. This snapshot can be checked in version control and shared with your team. To generate the snapshot, run

bash
npx directus schema snapshot ./snapshot.yaml
npx directus schema snapshot ./snapshot.yaml

To run non-interactively (e.g. when running in a CI/CD workflow), run

bash
npx directus schema snapshot --yes ./snapshot.yaml
npx directus schema snapshot --yes ./snapshot.yaml

Note, that this will force overwrite existing snapshot files.

Date-based snapshots

To keep multiple snapshot organized by date, create a folder snapshots in your project root directory add the following custom script to your package.json:

bash
"create-snapshot": "npx directus schema snapshot ./snapshots/\"$(date \"+%F\")\"-snapshot-\"$(date \"+%s\")\".yaml"
"create-snapshot": "npx directus schema snapshot ./snapshots/\"$(date \"+%F\")\"-snapshot-\"$(date \"+%s\")\".yaml"

When you run the command via npm run create-snapshot it will create a new snapshot with the following naming schema: [YYYY-MM-DD]-snapshot-[timestamp].yaml. This command can be run e.g by your deployment pipeline before each deploy on your server to keep a schema backup.

Applying a Snapshot

To make a different instance up to date with the latest changes in your data model, you can apply the snapshot. By applying the snapshot, Directus will auto-detect the changes required to make the current instance up to date with the proposed data model in the snapshot file, and will run the required migrations to the database to make it match the snapshot.

To apply the generated snapshot, run

bash
npx directus schema apply ./path/to/snapshot.yaml
npx directus schema apply ./path/to/snapshot.yaml

To run non-interactively (e.g. when running in a CI/CD workflow), run

bash
npx directus schema apply --yes ./path/to/snapshot.yaml
npx directus schema apply --yes ./path/to/snapshot.yaml

To diff the schema and database and print out the planned changes, run

bash
npx directus schema apply --dry-run ./path/to/snapshot.yaml
npx directus schema apply --dry-run ./path/to/snapshot.yaml

Creating Users

To create a new user with a specific role, run

bash
npx directus users create --email <user-email> --password <password> --role <role-uuid>
npx directus users create --email <user-email> --password <password> --role <role-uuid>

Updating User Password

To update the password of an existing user, run

bash
npx directus users passwd --email <user-email> --password <new-password>
npx directus users passwd --email <user-email> --password <new-password>

Creating Roles

To create a new role, run

bash
npx directus roles create --role <role-name>
npx directus roles create --role <role-name>

These roles are created with the minimum permissions required to properly access the App by default.

To create a new role with admin access, set the --admin flag to true, such as

bash
npx directus roles create --role <role-name> --admin true
npx directus roles create --role <role-name> --admin true

Count Items in a Collection

To count the amount of items in a given collection, run

bash
npx directus count <collection-name>
npx directus count <collection-name>

Client

For the client-side CLI, all functionality can be accessed by running npx directusctl <command>. You can also install @directus/cli on your project dependencies or globally on your machine. Note that if you run directusctl (installed globally) in a folder containing a project that has a version of @directus/cli installed, the running global CLI will forward it's execution to the local installed version instead.

Help & Documentation

The documentation for all commands can be accessed through the CLI itself. You can list all the available commands through directusctl --help command. If you want help for a specific command you can use directusctl <command> --help instead.

Instances

Most client-side CLI commands needs a running Directus instance in order to work. To connect the CLI to an instance, you can use directusctl instance connect command. These instance's configs are going to be saved on ~/.directus folder.

To manage the connected instances, you can use directusctl instance <command> commands.

Selecting instances

By default, commands will try using an instance named default when executing commands.

If you want to change which instance you want to use, either pass --instance <name> to the command, or configure instance variable on your project's Directus config file.

For example:

.directus.yml

yaml
instance: my-project
instance: my-project

I/O

The CLI is designed with ease of use and automation in mind, this means that you can change the way the output is made by setting how you want the data to be written to the terminal. We currently support three formats, table (the default one), json and yaml.

This makes it easier to parse and use data from Directus with other tools like jq, yq, grep or any other tools that accepts data from stdin

It's also worth mentioning that everything is data. Try for example running directusctl --help --format=json.

Table

The default output format. This is the "pretty" output, you'll most likely want to use this if you're not dealing with data in a way you need to pipe it to another command and/or store it for parsing.

This output will output colors and highlight content if it detects you're running in TTL.

JSON

This format will output JSON notation strings to your terminal. By default if TTY is detected, it will highlight (can be turned off with special flags) and prettify the output to make it easier to read.

Useful when you need to parse data using tools like jq for example.

YAML

This format will output YAML strings to your terminal. By default if TTY is detected, it will highlight (can be turned off with special flags) and prettify the output to make it easier to read.

Useful when you need to parse data using tools like jq for example.