TerminalAssistantChat

This project is a chatbot that uses OpenAI's GPT-3/4 model for generating responses. The model is capable of handling function calling and executing tasks based on instructions provided in the conversation. The chatbot can be run as a console-based chat application or as a Telegram bot.

Features

Execute bash code or terminal command: This allows the bot to execute bash commands and return the results directly in the chat.
Execute Python code or expressions: The bot can execute Python expressions and return the result. This can be helpful for quick calculations or other operations.
- Can install missing Python modules: If the bot requires a specific Python module that is not currently installed, it can handle the installation process.
Sends a file to/from a Telegram chat: The bot can send and receive files to/from a Telegram chat, which can be useful for sharing resources.
Gets the content of a website: The bot can fetch and display the content of a website or URI.
Can search queries using the ddgr command: The bot can search the web (using Google) with the bash command: ddgr --json -n 3 [search keywords].

Getting Started

Prerequisites

Python 3.10 or higher
OpenAI Python package
Telegram Python package
Docker

Environment Variables

Before running the application, you need to set the following environment variables:

OPENAI_API_KEY: OpenAI API key
TG_API_KEY: Telegram API key
TG_USER_ACCESS: A comma-separated list of authorized user IDs
OPENAI_MODEL: (Optional) A OpenAI model, default gpt-3.5-turbo-0613

Usage

Running as a Console-based Chat

To run the application as a console-based chat, execute the script with no arguments:

python main.py

Running as a Telegram Bot

To run the application as a Telegram bot, execute the script with the "telegram" argument:

python main.py telegram

Running the app in a Docker Container

Before you run the Docker container, you need to ensure that you have set the necessary environment variables in your .env file:

OPENAI_API_KEY: OpenAI API key
TG_API_KEY: Telegram API key
TG_USER_ACCESS: A comma-separated list of authorized user IDs
OPENAI_MODEL: (Optional) A OpenAI model, default gpt-3.5-turbo-0613

Once the environment variables are set, you can run the application inside a Docker container using the following command:

docker-compose up -d

If you want to run the script as Console-based Chat inside the Docker container:

docker-compose run bot python3 main.py

License

This project is licensed under the MIT License.

Acknowledgments

OpenAI for the GPT-3/4 model
Python-Telegram-Bot developers for the Telegram Bot wrapper

sstrelnikov/terminal-assistant-chat