As a non-IDE option, I offer this tool that I built for providing code context to (primarily) Claude Projects. But it works with OpenAI GPTs as well as LLM chats that don’t have “knowledge” files. It might be particularly useful with o1/o1-mini ATM, since those chats don’t yet allow attachments.
I basically paste my entire codebase in one operation into “Knowledge” in a Claude 3.5 Sonnet Project and work on each task in a single chat. It’s shocking how effective it is.
I’m not knocking Aider/Cursor/Continue, they’re all wonderful projects. It’s just that I find I’m more productive in a chat environment where I go through many iterations on AI generated code before I’m happy with the result. I also like knowing exactly what’s in the context and what the prompt is.
And yes, the tool was dogfooded / bootstrapped with itself 
https://github.com/cyberchitta/llm-context.py
Happy to get feedback from any users. Since this isn’t an elixir project, it might be better to provide specific feedback on the project issues tab, rather than here.