taskGPT

taskGPT is a Python-based command-line tool for running AI agents powered by OpenAI and Gemini models. It offers a guided setup experience, installs dependencies automatically, and configures API keys globally for seamless future use.

Features

• Run GPT-based agents from your terminal with a simple command
• One-time setup for dependencies and API keys
• Supports both OpenAI and Gemini API keys
• Persists API keys to system environment (Windows Registry / shell config)
• Automatically detects setup and avoids redundant installations
• Docker support for isolated environments
• Installable via PyPI or GitHub

Installation

Option 1: Install via pip

You can install taskGPT from either PyPI or GitHub:

From PyPI

pip install taskgpt

From GitHub

pip install git+https://github.com/KetanHegde/taskGPT.git

Option 2: Run using Docker

If you prefer using containers, taskGPT works great inside Docker.

Build the Docker image

docker build -t taskgpt .

Run the container

Unix/Linux

docker run -it --rm \
  -v "$PWD":/app \
  -w /app \
  -e OPENAI_API_KEY=your-openai-key \
  -e GEMINI_API_KEY=your-gemini-key \
  taskgpt

Windows CMD

docker run -it --rm ^
  -v "%cd%":/app ^
  -w /app ^
  -e OPENAI_API_KEY=your-openai-key ^
  -e GEMINI_API_KEY=your-gemini-key ^
  taskgpt

Windows Powershell

docker run -it --rm `
  -v "${PWD}:/app" `
  -w /app `
  -e OPENAI_API_KEY=your-openai-key `
  -e GEMINI_API_KEY=your-gemini-key `
  taskgpt

Usage

Once installed, run the tool using:

taskgpt

First Run

On the first launch, taskgpt will:

1. Install missing Python dependencies
2. Prompt you to enter GEMINI_API_KEY and/or OPENAI_API_KEY if not already set
3. Save them permanently in your system environment
4. Ask you to restart your terminal (only on first setup)

Subsequent runs will skip setup and launch the agent directly.

Development

To contribute to this project:

Clone the repository

git clone https://github.com/KetanHegde/taskGPT.git
cd taskGPT

Install dependencies

pip install -e .