This project provides an English-language shell for any operating system, powered by Large Language Models (LLMs). It allows users to execute complex commands and tasks using natural language, with features for iterative debugging and automatic package installation, making it suitable for developers and power users seeking a more intuitive command-line experience.

How It Works

Engshell leverages LLMs to interpret natural language commands and generate executable code, primarily Python. The core loop involves receiving user input, constructing a prompt for the LLM (with modes for text, code, or installation), executing the LLM's output, and handling errors by debugging or installing missing dependencies. This approach abstracts away complex syntax, enabling users to interact with their system through conversational prompts.

Quick Start & Requirements

• Install requirements: pip install -r requirements.txt
• Set environment variable: OPENAI_API_KEY or OPENROUTER_API_KEY.
• Run: python engshell.py
• Additional API key: UNSPLASH_API_KEY for image generation.
• See examples and usage: README

Highlighted Details

• Iteratively debugs generated code and automatically installs missing packages.
• Supports DALL-E for image generation and Unsplash API for wallpapers.
• Can perform complex tasks like data merging, plotting, and solving differential equations with SymPy.
• Includes safety warnings regarding arbitrary code execution and LLM unpredictability.

Maintenance & Community

No specific contributors, sponsorships, or community links (Discord/Slack) are mentioned in the README.

Licensing & Compatibility

The README does not explicitly state a license.

Limitations & Caveats

The project explicitly warns of potential arbitrary code execution, unintended consequences, and security vulnerabilities due to the unpredictable nature of LLMs. It strongly recommends running the software in a sandboxed environment.