ChatGPT CLI

A command-line interface (CLI) for interacting with OpenAI's ChatGPT using the chatgpt_rs library.

Features

• Send messages and receive responses from ChatGPT in a conversational manner.
• Save, load, and manage conversation history.
• Supports various commands for managing conversations, such as flush, save, load, and clear.
• Ask GPT to work with files via the CLI

Requirements

• Rust 1.56.0 or later
• An OpenAI API key

Installation

1. Clone the repository:

   git clone https://github.com/your_username/chatgpt_cli.git
   

2. Change to the project directory:

   cd chatgpt_cli
   

3. Compile the project:

   cargo build --release
   

4. (Optional) Add the compiled binary to your system's PATH for easier access.

Usage

To run the ChatGPT CLI, execute the following command, replacing your_api_key with your OpenAI API key:

./target/release/chatgpt_cli your_api_key

You can also pass your prompt directly after the api key

./target/release/chatgpt_cli your_api_key write a haiku

You can interact with ChatGPT by typing messages directly in the CLI. The following commands are also available:

• help: Displays help message
• flush: Clears the current conversation
• save [name]: Saves the current conversation under the specified name
• remove [name]: Removes a saved conversation
• load [name]: Loads a saved conversation
• clear: Clears all saved conversations
• list: Lists all saved conversations
• --file=[file(s)] --batch [optional, disable streaming process file(s) in batches] [message]: Sends a message related to a file that is also uploaded
• [message]: Sends a message

Flags:

• --gpt4: Use this flag to enable GPT-4 model
• --gpt35: Use this flag to enable GPT-3.5 model

License

This project is licensed under the MIT License.