Skip to content

awslabs/aws-shell

aws-shell - The interactive productivity booster for the AWS CLI

https://aws-developer-blog-media.s3-us-west-2.amazonaws.com/cli/Super-Charge-Your-AWS-Command-Line-Experience-with-aws-shell/aws-shell-final.gif

Installation

The aws-shell requires python and pip to install. You can install the aws-shell using pip:

$ pip install aws-shell

If you are not installing into a virtualenv you can run:

$ sudo pip install aws-shell

Mac OS X (10.11 El Capitan) users: There is a known issue with Apple and its included python package dependencies (more info at pypa/pip#3165). We are investigating ways to fix this issue but in the meantime, to install the aws-shell, you can run: sudo pip install aws-shell --upgrade --ignore-installed six

Once you've installed the aws-shell, you can now run:

$ aws-shell

To exit the shell, press Ctrl-D.

Upgrading the aws-shell

If you want to upgrade to the latest version of the aws-shell, you can run:

$ pip install --upgrade aws-shell

You can also use this upgrade command whenever a new version of the AWS CLI is released that includes new services and API updates. You will then be able to use these new services and API updates in the aws-shell.

Supported Python Versions

The aws-shell works on the same python versions supported by the AWS CLI:

  • 2.6.5 and greater
  • 2.7.x and greater
  • 3.3.x and greater
  • 3.4.x and greater

Configuration

The aws-shell uses the same configuration settings as the AWS CLI. If you've never used the AWS CLI before, the easiest way to get started is to run the configure command:

$ aws-shell
aws> configure
AWS Access Key ID [None]: your-access-key-id
AWS Secret Access Key [None]: your-secret-access-key
Default region name [None]: region-to-use (e.g us-west-2, us-west-1, etc).
Default output format [None]:
aws>

For more information about configure settings, see the AWS CLI Getting Started Guide.

Basic Usage

The aws-shell accepts the same commands as the AWS CLI, except you don't need to provide the aws prefix. For example, here are a few commands you can try:

$ aws-shell
aws> ec2 describe-regions
{
    "Regions": [
        {
            "Endpoint": "ec2.eu-west-1.amazonaws.com",
            "RegionName": "eu-west-1"
        },
        ...
aws> s3 ls
2015-12-07 15:03:34 bucket1
2015-12-07 15:03:34 bucket2
aws> dynamodb list-tables --output text
TABLENAMES     First
TABLENAMES     Second
TABLENAMES     Third

Profiles

The aws-shell supports AWS CLI profiles. You have two options to use profiles. First, you can provide a profile when you start the aws-shell:

$ aws-shell --profile prod
aws>

When you do this all the server side completion as well as CLI commands you run will automatically use the prod profile.

You can also change the current profile while you're in the aws-shell:

$ aws-shell
aws> .profile demo
Current shell profile changed to: demo

You can also check what profile you've configured in the aws-shell using:

aws> .profile
Current shell profile: demo

After changing your profile using the .profile dot command, all server side completion as well as CLI commands will automatically use the new profile you've configured.

Features

Auto Completion of Commands and Options

The aws-shell provides auto completion of commands and options as you type.

https://cloud.githubusercontent.com/assets/368057/11824078/784a613e-a32c-11e5-8ac5-f1d1873cc643.png

Shorthand Auto Completion

The aws-shell can also fill in an example of the shorthand syntax used for various AWS CLI options:

https://cloud.githubusercontent.com/assets/368057/11823453/e95d85da-a328-11e5-8b8d-67566eccf9e3.png

Server Side Auto Completion

The aws-shell also leverages boto3, the AWS SDK for Python, to auto complete server side resources such as Amazon EC2 instance Ids, Amazon Dynamodb table names, AWS IAM user names, Amazon S3 bucket names, etc.

This feature is under active development. The list of supported resources continues to grow.

https://cloud.githubusercontent.com/assets/368057/11824022/3648b4fc-a32c-11e5-8e18-92f028eb1cee.png

Fuzzy Searching

Every auto completion value supports fuzzy searching. This enables you to specify the commands, options, and values you want to run with even less typing. You can try typing:

  • The first letter of each sub word: ec2 describe-reserved-instances-offerings -> ec2 drio
  • A little bit of each word: ec2 describe-instances -> ec2 descinst
  • Any part of the command: dynamodb table -> Offers all commands that contain the subsequence table.

https://cloud.githubusercontent.com/assets/368057/11823996/18e69d16-a32c-11e5-80a2-defbaa6a8a80.png

Inline Documentation

The aws-shell will automatically pull up documentation as you type commands. It will show inline documentation for CLI options. There is also a separate documentation panel that will show documentation for the current command or option you are typing. Pressing F9 will toggle focus to the documentation panel allowing you to navigate it using your selected keybindings.

https://cloud.githubusercontent.com/assets/368057/11823320/36ae9b04-a328-11e5-9661-81abfc0afe5a.png

Fish-Style Auto Suggestions

The aws-shell supports Fish-style auto-suggestions. Use the right arrow key to complete a suggestion.

https://cloud.githubusercontent.com/assets/368057/11822961/4bceff94-a326-11e5-87fa-c664e1e82be4.png

Command History

The aws-shell records the commands you run and writes them to ~/.aws/shell/history. You can use the up and down arrow keys to scroll through your history.

https://cloud.githubusercontent.com/assets/368057/11823211/b5851e9a-a327-11e5-877f-687dc1f90e27.png

Toolbar Options

The aws-shell has a bottom toolbar that provides several options:

  • F2 toggles between fuzzy and substring matching
  • F3 toggles between VI and Emacs key bindings
  • F4 toggles between single and multi column auto completions
  • F5 shows and hides the help documentation pane
  • F9 toggles focus between the cli and documentation pane
  • F10 or Ctrl-D exits the aws-shell

As you toggle options in the toolbar, your preferences are persisted to the ~/.aws/shell/awsshellrc file so that the next time you run the aws-shell, your preferences will be restored.

https://cloud.githubusercontent.com/assets/368057/11823907/8c3f1e60-a32b-11e5-9f99-fe504ea0a5dc.png

Dot Commands

The aws-shell provides additional commands specific to the aws-shell. The commands are available by adding the . prefix before a command.

Exiting the Shell

You can run the .exit or .quit commands to exit the shell.

Creating Shell Scripts with .edit

There are times when you may want to take a sequence of commands you've run in the aws-shell and combine them into a shell script. In addition to the command history that's persisted to the history file, the aws-shell also keeps track of all the commands you've run since you first started your aws-shell session.

You can run the .edit command to open all these commands in an editor. The aws-shell will use the EDITOR environment variable before defaulting to notepad on Windows and vi on other platforms.

aws> ec2 describe-instances
aws> dynamodb list-tables
aws> .edit

Changing Profiles with .profile

You can change the current AWS CLI profile used by the aws-shell by using the .profile dot command. If you run the .profile command with no arguments, the currently configured shell profile will be printed.

aws> .profile demo
Current shell profile changed to: demo
aws> .profile
Current shell profile: demo

.cd

You can change the current working directory of the aws-shell by using the .cd command:

aws> !pwd
/usr
aws> .cd /tmp
aws> !pwd
/tmp

Executing Shell Commands

The aws-shell integrates with other commands in several ways. First, you can pipe AWS CLI commands to other processes as well as redirect output to a file:

aws> dynamodb list-tables --output text | head -n 1
TABLENAMES     First
aws> dynamodb list-tables --output text > /tmp/foo.txt

Second, if you want to run a shell command rather than an AWS CLI command, you can add the ! prefix to your command:

aws> !ls /tmp/
foo.txt                                    bar.txt

Developer Preview Status

The aws-shell is currently in developer preview. We welcome feedback, feature requests, and bug reports. There may be backwards incompatible changes made in order to respond to customer feedback as we continue to iterate on the aws-shell.

More Information

Below are miscellaneous links for more information: