gpt-pilot - Enhance Application Development with AI Automation

Introduction to GPT Pilot

GPT Pilot is an innovative project that explores the capabilities of Large Language Models (LLMs) in generating fully functional, production-ready applications. The unique aspect of GPT Pilot is its ability to automate a significant portion of code generation, potentially handling up to 95% of the coding workload. Despite this high level of automation, a developer's involvement is essential for the final 5% of the project until the advent of full artificial general intelligence (AGI).

Key Features

Code Generation and Application Building: GPT Pilot goes beyond simple code generation; it builds entire applications. This ambitious approach aims to streamline the development process significantly, relying on AI to write most of the code.
Research-Oriented: The project serves as a research initiative evaluating how effectively LLMs can be used in practical application development. Through this research, GPT Pilot seeks to uncover insights that could shape the future of AI-assisted programming.
Collaborative Development: Developers play a critical role in overseeing the implementation process, ensuring that while AI handles the bulk of the workload, human oversight maintains quality and adherence to project requirements.

Getting Started with GPT Pilot

Requirements:

Python 3.9 or higher.

Setup Instructions:

Clone the Repository: Start by cloning the GPT Pilot repository using Git.
```
git clone https://github.com/Pythagora-io/gpt-pilot.git
cd gpt-pilot
```

Set Up Environment: Create and activate a Python virtual environment.

python3 -m venv venv
source venv/bin/activate  # For Windows use `venv\Scripts\activate`

Install Dependencies: Utilize pip to install all the necessary packages.
```
pip install -r requirements.txt
```
Configure Settings: Copy the example configuration and adjust it.
```
cp example-config.json config.json
```
Update this file to include your API key, database settings, and any paths to be ignored by the AI system.
Start GPT Pilot: Finally, initiate the main program.
```
python main.py
```

The generated code will be saved in a 'workspace' folder tailored to your application.

Examples and Usage

Examples of applications built with GPT Pilot can be found in its wiki page. These examples illustrate the diverse applications and capabilities of GPT Pilot.

Docker Setup

For those preferring Docker, GPT Pilot can be started using Docker containers. Follow these steps:

Clone the repository as instructed above.
Modify the docker-compose.yml to suit your environment requirements.
Build the Docker image with:
```
docker compose build
```
Launch the containers:
```
docker compose up
```
Access the web terminal to start GPT Pilot with:
```
python main.py
```

PostgreSQL Support

While SQLite is the default database, PostgreSQL is also supported. To switch to PostgreSQL, install additional packages and update the config.json with your database URL.

Command Line Interface

GPT Pilot provides several CLI commands for managing projects:

List Projects:
```
python main.py --list
```
Continue from Last Step:
```
python main.py --project <app_id>
```

Continue from Specific Step:

python main.py --project <app_id> --step <step>

In summary, GPT Pilot is a future-oriented tool designed to simplify and revolutionize application development through AI automation, while still recognizing the indispensable role of human oversight in the process.