After the agent makes a change I have a rule that the agent must run my program so it can monitor the terminal output. I’m running a flask application so it will just do python main.py in the terminal. When it tries to run the command the terminal doesn’t do anything. It just spins like its waiting and will stay there until I stop it. I notice in new composers it will work the first few times but then it will stop working. I will tell the ai this is happening then it might fix it or not. But if it does fix it, it will only fix that time and will stop working the next time.
I’ve probably used over 100 credits in the past few days just by telling composer that its not showing anything in the terminal. And I cant just keep creating new composers, its inefficient.
@Oliver_Michalik you are reviving here a thread thats from February and March about a totally different Cursor version that had a different usage of Terminal.
Its good to help others with the same issue find solutions. But this was not the same issue. I would appreciate if you really stay focused on same Cursor version.
I have been using cursor for over a year and I can confidently say that this is a big issue for me and it has not gotten fixed ever. The only “hack” I have found is to open a new chat but even that only lasts a little. It freezes, never completes or totally ignores the terminal that the agent opened. I really hope this gets fixed in a version soon because it would help with a lof of debugging.
