WTF - Natural Language to Shell Commands

WTF is a command-line tool that converts natural language descriptions into shell commands using AI. Just describe what you want to do, and WTF will generate the appropriate shell command.

Installation

Install from PyPI: (not yet available)

for now, you can install from the source code:

pip install -e .

Configuration

Before using WTF, you'll need to configure at least one AI provider if OPENAI_API_KEY or ANTHROPIC_API_KEY is not set in your environment. WTF supports OpenAI and Anthropic.

Set up your preferred provider's API key:

wtf-config set-key openai sk-your-api-key-here
wtf-config set-key anthropic sk-ant-your-api-key-here

Set your default provider (optional):

wtf-config set-default openai

if you need to see what the config file looks like, you can use the following command:

wtf --show-config

all configuration is stored in ~/.config/wtf/config.yaml. You can edit this file directly if needed.

Usage

Basic usage:

wtf find all PDF files modified in the last 7 days

Specify a different provider:

wtf -p anthropic list all docker containers including stopped ones

Use a specific model:

wtf -m gpt-4 create a tar archive of all jpg files

Execute the command directly:

wtf -e list all empty directories nested in current directory

Show debug information:

wtf -d "find largest files in current directory"

Show history:

wtf --history

Supported Providers and Models

• OpenAI

  • gpt-3.5-turbo
  • gpt-4
  • gpt-4o (default)

• Anthropic

  • claude-3-sonnet
  • claude-3-opus
  • claude-3-haiku
  • claude-3-5-sonnet (default)
  • claude-3-5-haiku
  • claude-3-5-opus

Testing

Install test dependencies and package

pip install -e ".[test]"

Run all tests (including integration tests)

pytest -v

Run only unit tests (skip integration tests that need API keys)

pytest -v -m "not integration"

Run with coverage report

pytest -v --cov=wtf

Run with coverage and generate HTML report

pytest -v --cov=wtf --cov-report=html

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.