gpt-engineer

README

GPT-Engineer

GPT-Engineer is a versatile software engineering tool built with Go. It offers a host of features designed to streamline your development workflow.

Features

1. Code Generation: Automatically generate full codebases in any language for the project you describe.
2. Automated Testing: Tools for automatically testing code to catch bugs and other issues.
3. Performance Profiling: Features for measuring and improving the performance of your code.
4. Logging: All input and output with OpenAI is written out to files.
5. Prebuilt Executables: Executables available for most systems so no need to build from source code.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Download and Run

You can just download the executable from the Release section on GitHub.

1. Download the archive that matches your system here, https://github.com/go-pax/gpt-engineer/releases
2. Extract the contents of the archive into a folder where you will run gpt-engineer
3. Open a terminal/command prompt and cd to the folder with gpt-engineer
4. Generate an OpenAI API key at, https://platform.openai.com/account/api-keys and copy the key.
5. Set the OPENAI_API_KEY environment variable with the key in the terminal/command prompt,
   • (macos/linux) export OPENAI_API_KEY=sk_123456789
   • (windows) set OPENAI_API_KEY=sk_123456789
6. Run gpt-engineer --help to view the arguments you can set
   • (macos/linux) ./gpt-engineer --help
   • (windows) gpt-engineer.exe --help
7. Run gpt-engineer and create your first project
   • (macos/linux) ./gpt-engineer -prompt "hello world app in go" ./output_folder
   • (windows) gpt-engineer.exe -prompt "hello world app in go" output_folder
8. (optional) Choose to execute the source code by typing "yes" or "no" when prompted. This requires you have everything installed to execute the source code.

Developer Setup

Prerequisites

What things you need to install the software and how to install them:

go get -u github.com/go-pax/gpt-engineer

Installation

A step-by-step series of examples that tell you how to get a development environment running:

go install github.com/go-pax/gpt-engineer

Usage

Here are the steps to setup your GPT-Engineer environment. You can use both OpenAI or Azure OpenAI.

Setup:

Get go packages,

go mod tidy

OpenAI

Run using OpenAI API key. Set the following environment variable:

export OPENAI_API_KEY=[your openai api key]

Attempts to use GPT4 if not then fallback to GPT-3.5.

Azure OpenAI

You will need an Azure account and an OAI resource, https://oai.azure.com/portal

export OPENAI_API_KEY=[your azure openai api key]
export OPENAI_API_BASE=https://[deployment name].openai.azure.com/

Run:

• Create an empty folder. If inside the repo, you can run:
  • cp -r projects/example/ projects/my-new-project
• Edit main_prompt file in your new folder to specify what you want to build
• Run go run . ./projects/my-new-project, default language is English, if you want to use other language, and -lang argument like go run . -lang=Chinese
  • (Note, go run . --help lets you see all available options. For example --steps use_feedback lets you improve/fix code in a project)

(optional) Database You should use the scheme of the database instead of the path to your projects. View the readme files in the database folder to understand how to use. example, instead of ./projects/my-new-project use file://./projects/my-new-project to ensure the file database is used.

(optional) Azure

• use the argument -model [your deployment name], when using Azure this must be set

Results:

• Check the generated files in projects/my-new-project

Fine Tuning:

You can specify the "identity" of the AI agent by editing the files in the identity folder. Editing the identity, and evolving the main_prompt, is currently how you make the agent remember things between projects.

Logging:

Each step in steps.go will have its communication history with GPT4 stored in the logs folder.

Contributing

Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.

License

This project is licensed under the MIT License - see the LICENSE.md file for details.

Thanks

Thanks to the projects;

• AntonOsika/gpt-engineer that created GPT-Engeneer
• geekr-dev/gpt-engineer this project is based off