ShellLM

A Versatile PowerShell-based LLM Interface

Overview

ShellLM is a feature-rich and flexible terminal application that allows users to have text-based conversations with large language models (LLMs), including GPT-4, Gemini, and Claude 3. It provides an intuitive command-based interface, enabling users to easily interact with the models and explore their capabilities.

Features

• Seamless communication with LLM APIs such as GPT-4, Gemini, and Claude 3
• Colorful formatted messages for easy readability
• Comprehensive conversation history and navigation
• Conversation saving and loading to/from JSON files
• Retry option to generate alternative responses
• AskLLM mode for concise answers to specific questions
• Supports many models from several providers:
  • OpenAI
    • GPT-3.5-turbo
    • GPT-4o
    • GPT-4o-mini ✨
  • Google
    • gemini-pro
    • gemini-1.5-pro ✨
    • gemini-1.5-flash ✨
  • Anthropic
    • claude-3-haiku
    • claude-3-sonnet
    • claude-3-opus
    • claude-3.5-sonnet ✨
• Image generation: Generate images using the DALL-E API (requires an OpenAI API key)
• Custom instructions: Set specific guidelines for the model to follow during conversations
• Automatically loads conversation 'autoload.json' on startup, automatically saves last conversation

Installation

1. Obtain an OpenAI API key:

   • Sign in to your OpenAI account and navigate to your account settings.
   • Select "API Keys" and create a new API key if you don't already have one.

2. Set the API key as an environment variable:

   • Open a Command Prompt and enter the following command:

     [System.Environment]::SetEnvironmentVariable("OPENAI_API_KEY", "sk-...", "User")
     

   • Replace "sk-..." with your actual API key.

3. Clone or download the repository:

   git clone https://github.com/Digit404/ShellLM.git
   

Usage

1. Start the ShellLM terminal interface

   .\ShellLM.ps1
   

2. Interact with the model

   • Type your messages and press Enter to send them to the model.
   • Use the available commands (listed below) to manage conversations, generate images, and customize the experience.

• Alternatively, you can append your prompt after the file name to get a quick response without entering the interactive shell:

  .\ShellLM.ps1 "What is the capital of Ecuador?"
  

Commands

• /bye (/goodbye): Exit the program with a farewell message.
• /help: Display a list of available commands and their descriptions.
• /exit (/quit): Exit the program immediately.
• /save [filename] (/export): Save the current conversation to a JSON file.
• /load [filename] (/import): Load a previous conversation from a JSON file.
• /hist (/list, /ls, /history): Display the conversation history.
• /back [number]: Go back a number of messages in the conversation.
• /retry: Generate another response to your last message.
• /reset (/clear): Reset the conversation to its initial state.
• /model [model]: Change the model used for generating responses.
• /imagine {prompt} (/generate, /image, /img): Generate an image based on a given prompt.
• /copy [number of messages back] (/cp): Copy the last response to the clipboard (without formatting).
• /paste [prompt] (/clipboard): Provide the model with the content of your clipboard as context.
• /rules [instructions] (/instructions): Set custom instructions for the model to follow during the conversation.
• /config: Change other persistent settings.

Troubleshooting

If you encounter an issue initially running the program, make sure you have enabled running scripts on your system by running this command powershell Set-ExecutionPolicy Bypass -Scope User

If you encounter any further issues, please double-check that you have correctly set the OPENAI_API_KEY environment variable. If the problem persists, please open an issue on the GitHub repository for further assistance.

Note: This script is still under development and may undergo further changes in the future.