Is Cursor aware of API responses?

General
1

So I’ve found myself incrementally working towards code that uses the OpenAI transcribe / translate API correctly. It’s a slog; lots of questions and delays and attempts.

I have another topic open about why Cursor seems to be ignoring the OpenAI documentation it has access to. I’m not sure how much of that is contributing to the slog.

This is a different question: Is Cursor automatically aware of the API response and can I just, like, @ mention the response or tell Cursor to reason about the response it got back?

For example, there’s an obvious difference between getting back a big complex json object vs getting back a single text string. The code for each case is different. It would be magical if Cursor can automatically include the response in its reasoning and just edit the code to correctly handle the response.

2

The way you can do this with Cursor is using jupyter notebooks. Printing the API response to the cell. Then Cursor can read the cell outputs and answer questions with that context.

3

I google for that. So setup multiple tools like Anaconda and Jupyter, and then a cell in a Jupyter notebook can act like a variable that can be referenced in a question?

Is there something with fewer steps? For example, it seems like a simpler process would be to just write the response to response.txt and then reference that file.

Or am I missing something fundamental?

4

that could also work

1 Like