Raw LLM Responses

Recently I have dev a new prompt . I find that if your having a long session. Even I can forget and make many miss turns and even forget to use a object oriented methodology or even pick the right one . I also leave projects and sessions and have to return to the complexed code and maybe even forget where I was . So these long sessions and many code changes and total development with a model can become " let's abandon this chat and cut and paste to new chat ... " Which also changes it's jab it's for this project .. So now . I ask the model to review the application on a podcast style . IE create a podcast for this project . Let's discuss what future services and features which may need to be added and which parts of the code could be for from refactoring without loosing functionality . What are the best approaches for this task and how could it be done better . Should this project be abandoned ? Is it a useful library or should it be a API ? How should we implement this app and which methodologies should we use ? Based on. This foresight how should we proceed ... Should we refactor or rebuild with these new suggestions and practices mind? The model will whistle and buzz ... A produce a very good project outline and discussion ( maybe a few episodes ) as well as a planning strategy ? Any identified bugs or concerns . With some suggested reactions or structural changes . A full upgrade or rebuild pathway . The basic implementation or full implementation. Of your app . Often it will convert many things in to the plugin pattern and decorator patterns and builder factory's. Etc . Making it easy to use later ! Or upgrade without damaging the core app . It often identifys this as the main cause or problems . So removing the core app . And componetozing all other services making your app more pmuginable .. Often it will praise the work or idea ,or berate the implementaion etc . The podcast it produces is really professional and discussions are amazing as it seems to simulate You ! And a architect or developer and a interviewer ( IE the architect is the AI / the developer is you . And the Interviewer is the reasoner ! ... The discussion, is also like a sprint in which it tells you where you are and where your focus should be . It generally fully upgrades and refactors your app very professionally ! It will also go online to discover other apps which are similar and implement potential future services optionally ... I mean ! Just ask it to discuss and create a podcast about your app ! .... Use that podcast to prompt the model ! And it produces spectacular work ! ... You will be super amazed ! I use qwen ! And it's changed the quality of qwens outputs with me in all my future sessions ... It was like I unlocked it !