Multiple issues and suggestion

  1. There is no option to choose fast or slow, gpt4 or gpt3.5 when using my own key
  2. Whenever I give it an extended command that needs to generate 20 plus file it gets stuck at Finding related files for XXXX/XXXXXX.js
  3. Minor but sometimes it leave the file type markdown at the top and bottom e.g ``` javascript

Is there a way to see what it is sending to OpenAI and “resume” if the it breaks somehow?
Additionally same thing for changing the GPT version to limit costs?
Is there a debug option to see what’s going on and what’s failing?

The debugging would allow me to see whats going on so we I can change, similarly if I am hitting any openAI limits ( less likely ), slow option should be able to mitigate that

  1. Fast and slow requests are just a thing by Cursor requests. If you use your own API key, you only have one speed.
  2. Could you elaborate? What do you mean by “20 plus file”?
  3. Yea, got that a few times as well. We need to adjust some things there.
  1. An option to limit the req/min in case someone is using a new account with lower limits
  2. A project that is generating more than 20 files, I tried the same prompt 4 times before and it got stuck at different files but at the same point i.e. Finding related files for XXXXXX.js
  3. Thank you

Additionally, there is a billing bug, where can I contact without it being public