Your best use cases for the new interpreter mode?

What are your first experiences with the new “Interpreter Mode”? It seems to me that this is a very powerful feature. Therefore, I would be interested in what you have used it for so far.

I’ll start with two use cases:

  1. Edit multiple files in the same way in one step. Prompt: Add the code comment “// Written by AI” to all files ending with “Service”. Please search the entire directory structure of the current project for this. Write an executable Python script for this!

  2. Generalize code and extract it into another file: Add two Code snippets as extra context and prompt: “Generalize these two snippets and extract it in a new file with name XXXService”

Let’s see what you have :slight_smile:


I’m also really curious to hear what folks are doing with interpreter mode…

…it seems powerful but right now I don’t really get it…


interpreter mode is getting a real boost soon!

1 Like

Has it already?

It’s a big weakness of Cursor, lot of people including me have no idea how most of the features works. Would be nice to have proper tutorial video or at least some documentation.


Is interpreter mode like the AI project feature that auto generates a complete project for you but based on a prompt?

If you don’t know this yet:

Sure, I’d be glad to point you to the documentation, but first, could you clarify which documentation you’re referring to? This will help me provide you with the most accurate information and direction.

Screenshot 2024-03-28 at 19.20.54

Nothing for interpreter there

Ah, okay, sry.

Interpreter basically creates a script based on the chat (depending on the problem) and lets you run the script directly from the chat.

1 Like

Does interpreter mode only work for Python? What are some use cases for interpreter mode?

no works for everything but you need the Python Extension and also Jupyter

I should have specified updated doc.

I"m using interpreter to process all kinds of data locally, such as CSVs, spreadsheets, etc. It uses Python, has access to basically the whole file-system and can execute any operation, so I usually use “auto-execute” (I’m inside a virtual machine though, so it’s relatively safe). It helps a lot and you don’t have to upload any data anywhere, like with ChatGPT’s Interpreter.

I’ve tried using it for batch editing codebase, but found it not as strong there, as it sometimes misedits files or saves to the wrong file for some reason. However, for simple batch edits in the codebase it usually works well.