I am a born again coder thanks to GPT-4. Used to code 20 years ago but have recently started to build code again, working with GPT-4 through the logic and building new applications. I also find Ghostwriter very useful although it is not as powerful as using GPT-4 itself in ChatGPT…shame with the limitations though!
Anyways I would like to see a feature where one can plug a Repl project and GPT-4 can provide a summary of what the code does, spots errors and provides suggestions or improvements to the logic and makes it more optimized. Does anyone have any ideas or ways I could do that?
Currently I am struggling with ChatGPT (GPT-4) as I have to provide context everytime with 8k tokens limit and the 25 messages every 3 hours. Have asked to increase to 32K tokens soon but I was wondering if there are any resources within the Replit community around such a functionality as described above?