Hello everyone, meet again, I'm GitHub Select!
Background
During software development, it is often necessary to write code according to requirements. However, writing code is a time-consuming and tedious task. The GPT Engineer project aims to solve this problem by generating a complete code base based on the user's needs. You simply specify what you want to build, and the AI will ask you for details and ultimately generate the corresponding code.
Today I would like to recommend a GitHub open source project AntonOsika/gpt-engineer, which has more than 36.4k stars on GitHub, and the introduction of the project in one sentence is: "Specify what you want it to build, the AI asks for clarification, and then builds it."
Project Introduction
The GPT Engineer project is designed to provide a tool that is easy to adapt, extend, and let your agents learn what your code looks like. It generates the entire codebase at the prompt.
Here is the design philosophy of the project:
It is briefly summarized as follows:
- Simple and easy to use
- Flexible and easy to add custom "AI steps", see steps.py
- Build the user experience progressively, including: 1. Advanced tip 2. Provide feedback to AI that will be remembered for a long time
- Rapid interaction between AI and humans
- Simplicity, all calculations can be interrupted and saved to the file system
Below is a demo video of the project: Downloads/gpt-engier.mp4
How to use
Select Stable or Development.
For stable releases:
- pip install gpt-engineer
For Development Edition:
- git clone https://github.com/AntonOsika/gpt-engineer.git
- cd gpt-engineer
- pip install -e . (Or: for virtual environments, you can run make install & source venv/bin/activate)
In addition, you need to perform the following steps
Using an OpenAI API key (preferably with GPT-4 access), run:
- export OPENAI_API_KEY=[your API key]
Alternative setup methods for Windows systems:
- Execute set OPENAI_API_KEY=[your API key] on the command line
- Execute $env:OPENAI_API_KEY="[your API key]" in PowerShell
run
- Create an empty folder. If inside the repository, you can run: cp -r projects/example/ projects/my-new-project
- Fill in the prompt file in the new folder
- Run gpt-engineer projects/my-new-project• (Note that you can see all available options with the gpt-engineer --help command.) For example, use --steps use_feedback to improve/fix code in your project)
outcome
Check the files generated in the projects/my-new-project/workspace folder.
Here's a Star trend chart for the project, which represents how active the project is:
For more project details, please see the link below.
Open source project address: https://github.com/AntonOsika/gpt-engineer
Open source project by Anton Osika
Here are all the members involved in the construction of the project:
Follow us to explore interesting open source projects.