Skip to main content

Shell Whiz: AI assistant right in your terminal

Project description

PyPI PyPI - Downloads GitHub stars GitHub commit activity GitHub license

Shell Whiz is an AI assistant for the command line. It will help you find the right command for your task.

All features

  • Command suggestions: It suggests shell commands based on your input. For example, if you want to know the timestamp of a file, you can run ?? what is the timestamp of file.txt and it will suggest stat -c %y file.txt, which will print the last modification time of the file.
  • Command explanations: It will try to explain, piece by piece, what the suggested command will do, so you can learn new things about your shell and the commands you use.
  • Revisions: If the suggested command is not exactly what you need, you are on a different platform, or you want to see other options, you can ask for a revision. It will suggest a different command that is similar to the previous one.
  • Customization: You can customize the behavior of Shell Whiz by using command line arguments. For example, you can set PowerShell as your shell, disable automatic explanations, or use the gpt-4 model instead of the default gpt-3.5-turbo model.

Installation and setup

To install Shell Whiz, run the following command:

pip install shell-whiz

This will add the sw command to your PATH.

To use Shell Whiz, you need an API key from OpenAI. You can obtain this key by visiting https://platform.openai.com/account/api-keys. Once you have the key, you can configure Shell Whiz by running the following command:

sw config

Free API plan for new users

New users receive $5 for free to try and test the API during the first 3 months. However, it is recommended to upgrade to a paid plan in order to have a more comfortable experience using Shell Whiz. This is because the free plan has restrictions on the number of requests allowed per minute.

Upgrading

To upgrade Shell Whiz, run the following command:

pip install --upgrade shell-whiz

Usage

You can run Shell Whiz directly using sw ask, but I recommend creating an alias for it. For example, you can add the following line to your ~/.bashrc file:

alias ??='sw ask'

To track API usage and costs, you can check the OpenAI API Usage page.

Tips

  • If you want to pass an argument that starts with a hyphen, you can use -- to separate the command from the arguments. For example, sw ask -- emulate ARM kernel on versatilepb architecture -cpu cortex-a8.
  • Add -m gpt-4 or --model gpt-4 to use the gpt-4 model instead of the standard gpt-3.5-turbo model. However, this will cost more and may take longer.
  • Add --explain-using-gpt-4 to use the gpt-4 model for the explanatory part.
  • Use -n or --dont-explain to disable automatic explanations. You can still request an explanation through the menu when a command is suggested.
  • Use --dont-warn to disable automatic warnings.
  • Pass -p "..." or --preferences "..." to set preferences for generating commands. This is most useful for setting the shell, but can be used to set any other preferences as well. For example, sw ask -p "I ussually use PowerShell" will generate commands that work in PowerShell. By default, this parameter is set to I ussually use Bash on Linux.

The original author of the program usually uses alias ??='sw ask --dont-warn -n --' because he has a good understanding of the command line and knows about dangerous commands. However, if he doesn't understand the generated command he chooses to explain it using GPT-4 via the menu.

You can choose the settings that work best for you.

More examples

License

Shell Whiz is licensed under the GNU General Public License v3.0.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

shell_whiz-1.2.1.tar.gz (22.1 kB view hashes)

Uploaded Source

Built Distribution

shell_whiz-1.2.1-py3-none-any.whl (23.0 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page