apna-college/SigmaGPT — reverse-engineered prompt
Reverse engineered prompt
Build me a full stack ChatGPT style web app called SigmaGPT using the MERN stack and the OpenAI API. I want a simple chat screen where a user can type a message, send it, see the assistant response come back in the conversation, and keep the experience feeling close to ChatGPT with a clean layout, loading state, and readable message bubbles.
Please set up the frontend and backend so they run separately, with the backend safely handling the OpenAI key through environment variables instead of putting secrets in the browser. Use MongoDB where it makes sense for storing app data like chats or messages, but keep it simple and easy to run locally. Add basic error handling for empty prompts, API failures, and server issues.
Make the project ready to start with clear install and run instructions. If OpenAI’s API has changed, look up the current docs online and use the right approach.
Want more depth? Deep Reverse