Command line interface (CLI)

See how to use and manage your projects directly from your terminal. Anything you can do within the Console can be done with the CLI.

The CLI uses the git interface and the REST API to accomplish tasks. Its source code is hosted on GitHub.

1. Install 

To install the CLI, use either Homebrew (on Linux, macOS, or the Windows Subsystem for Linux) or Scoop (on Windows):

$ brew install platformsh/tap/platformsh-cli
$ scoop bucket add platformsh
$ scoop install platform

For manual installation, you can also download the latest binaries.

Upgrade using the same tool:

$ brew upgrade platformsh-cli
$ scoop update platform

2. Authenticate 

To list and manage your projects, authenticate by running the following command:


This process opens a browser tab for you to log in. It also creates certificates on your computer for SSH.

Once you are logged in, a list of your projects appears, along with some tips for getting started. If you experience authentication issues or want to force a login, run the command platform login.

3. Use 

Now you can run actions on your projects such as branching and merging. You can also simulate a local build of your codebase as if you were pushing a change to, including your services and data.

Get a list of all available commands with:

platform list

To get more information on a specific command, preface it with help:

$ platform help get
Command: project:get
Aliases: get
Description: Clone a project locally

 platform get [-e|--environment ENVIRONMENT] [--depth DEPTH] [--build] [-p|--project PROJECT] [--host HOST] [-i|--identity-file IDENTITY-FILE] [--] [<project>] [<directory>]

  project                            The project ID
  directory                          The directory to clone to. Defaults to the project title

  -e, --environment=ENVIRONMENT      The environment ID to clone. Defaults to the project default, or the first available
      --depth=DEPTH                  Create a shallow clone: limit the number of commits in the history
      --build                        Build the project after cloning
  -p, --project=PROJECT              The project ID or URL
      --host=HOST                    The project's API hostname
  -i, --identity-file=IDENTITY-FILE  An SSH identity (private key) to use
  -h, --help                         Display this help message
  -q, --quiet                        Do not output any message
  -V, --version                      Display this application version
  -y, --yes                          Answer "yes" to any yes/no questions; disable interaction
  -n, --no                           Answer "no" to any yes/no questions; disable interaction
  -v|vv|vvv, --verbose               Increase the verbosity of messages

 Clone the project "abc123" into the directory "my-project":
   platform get abc123 my-project

Select the right project and environment 

When you are in an empty directory or a directory not associated with a specific project, if you run a command that requires a specific project and environment, you are prompted to select them.

$ platform environment:info
Enter a number to choose a project:
  [0] My project (xb3pfo734qxbeg)
  [1] A great project (3p5fmol45kxp6)
  [2] An even better project (rjify4y564xaa)

If your working directory is inside a local checkout of your project repository, your project and environment are detected automatically.

You can always specify the project and environment in two ways:

  • As arguments for the command:

    $ platform environment:info --project=my-project --environment=staging
  • With environment variables:

    export PLATFORM_PROJECT=my-project;
    export PLATFORM_BRANCH=staging;
    platform environment:info

In multi-app projects, this applies also to selecting the right app (the environment variable would be PLATFORM_APPLICATION_NAME).


If you check out a project via Git directly and not using the platform get command, the CLI may be unable to determine what project it’s in. You might run a CLI command from within a project directory you’ve checked out and get an error like this:

[RootNotFoundException] Project root not found. This can only be run from inside a project directory.

Then the CLI hasn’t been able to determine the project to use. To fix this, run:

platform project:set-remote <PROJECT_ID>

Replace <PROJECT_ID> with the ID of your project. You can find that in the Console or by running platform projects to list all accessible projects.

Choose between the CLI and Git commands 

Some CLI commands (especially many within the environment namespace) have some overlap with Git commands. Generally, they offer more options than the Git commands alone. For example, platform push offers options such as --activate (to active an environment before pushing) and --no-wait (so you can continue working without waiting for the push to complete).

For all of them, you don’t need to configure a Git remote. It’s enough to have a project ID.

An example of how this affects commands is that when you run platform merge, it doesn’t affect your local codebase. You don’t even need the code locally. The code is only merged between environments remotely.

Customize the CLI 

You can customize how the CLI operates and what it returns with a configuration file (.platform/local/project.yaml) or environment variables. For details, see the customization instructions on GitHub.

Automate repetitive tasks 

You might want to use the CLI in a script to automate repetitive tasks such as synchronizing your files locally. In such cases, you want to customize the CLI to bypass any confirmation questions. You can set the answer to every question as yes using the PLATFORMSH_CLI_NO_INTERACTION environment variable.

For instance, to locally sync every mount point for your app named app, you could use this command:

export PLATFORM_PROJECT=my-project;
export PLATFORM_BRANCH=main;
platform mount:download --all --app app --target local-backup

Autocomplete commands 

The CLI provides tab autocompletion for commands, options, and some values (your projects, valid regions).

Your system must include the bash-completion package or an equivalent. This isn’t available by default on macOS, but can be installed via brew. Check your home directory and ensure that the file ~/.platformsh/ is being included by your shell.

If you experience issues, run platform self:install to attempt a reinstall of this utility.

Run commands on your container 

You can use the CLI to run commands on your container. You can use any command you’ve added in dependencies or a hook.

The syntax looks like the following:

platform ssh -- <COMMAND> <ARGUMENTS>

For example, to run a specific Python script named on your current environment, run the following command:

platform ssh -- python

Or to use Drush to rebuild the cache on the feature environment, run this command:

platform ssh -e feature -- drush -y cache-rebuild