Cursor always assumes, without checking the actual code

I have noticed several times that Cursor assumes the actual implementation of code and provides incorrect answers that do not reflect our current codebase.

I constantly have to ask Cursor to check the real code implementation, which should be done by default. This issue has occurred multiple times, not just once. It seems that Cursor only has a very general idea of our code and responds based on that.

hi @idham thanks for your post. This occurs sometimes if an AI model thinks it has enough info from context.

Have you tried asking the agent to check existing code before proceeding?

This topic was automatically closed 22 days after the last reply. New replies are no longer allowed.