Cursor unable to correctly detect command completions

With me it started to occur with the last update.

Version: 1.3.9 (user setup)
VSCode Version: 1.99.3
Commit: 54c27320fab08c9f5dd5873f07fca101f7a3e070
Date: 2025-08-01T20:19:30.453Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.19045

I work on windows 10 using Drupal in WSL2. Cursor doesn’t seem to detect the end of a command anymore and keeps on spinning.

Same issue here. I really like how Terminal is implemented in the newer versions, but because of this bug, I have to re-install the older version (1.2.4) to get it working.

Version: 1.3.9 (user setup)
VSCode Version: 1.99.3
Commit: 54c27320fab08c9f5dd5873f07fca101f7a3e070
Date: 2025-08-01T20:19:30.453Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.26120

Running the development in WSL2.

Please fix the detection of the completed command.

1 Like

It still does not wait until “npm install” is finished

Can get around it by telling cursor to always use is_background_False for all tool calls. Best as a rule or memory until fixed. Only seems to be an issue with Cursor’s agent specifically, not external APIs.

This i seriously annoying. It’s just about every single time it tries to use the console. This needs to get fixed. If I cancel my paid subscription to cursor, it will be because of this issue.

If it matters, I am using remote SSH.

Version: 1.4.3 (user setup)
VSCode Version: 1.99.3
Commit: e50823e9ded15fddfd743c7122b4724130c25df0
Date: 2025-08-08T17:34:53.060Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.26100

1 Like

I’m in the same boat. The worst part is, I might kick off a whole bunch of commands before going to bed, and it got hung up on #2 of 10. When it happens while I’m at the machine, it’s annoying but solvable. IF I notice in time. Sometimes I try to fix it and I get “tool call timed out” or something.

Usually what i have to do (and this solves it 99% of the time, but ugh) is go to that terminal window and just hit enter.

Sitting here spinning on this:

image

I just hit enter there in the terminal and it’ll go again…
And it’ll sit here and spin…

And hit in the terminal window again. And again. And again. Seems like this, as an issue, would be one of the easiest to solve. Why doesn’t it just always have the terminal window output to a file it’s watching? In the case of multiple chat/agents, each likely has some unique ID which could be part of the file so its only monitor what it needs to watch.

Update: With 1.4.3 Terminal in WSL2 works for me 90% of the time. It’s a huge improvement, so thanks for fixing this.

How does one choose which type of terminal? Most of the time, pretty sure mine’s picking Powershell because it likes to run ps commands.

And it pretty much opens whatever terminals whenever it chooses.

I’m experiencing the same issue since the last update. When the agent completes a task, it temporarily freezes. After a while, Windows displays a message saying the task has stopped responding and asks whether to force close it or keep waiting. If I choose to keep waiting, the cursor freeze resolves after a short time.
I’m on Windows 11.

I’m still dealing with it daily. I’ve been looking at anything and everything within my experience level to solve it and I can’t find anything. Other than restarting Cursor, which is a 100% dumb solution. If I want to tell the AI to implement a feature, then run the quality checks (linter, typescript, test cases) and fix anything it finds… it’ll get stuck before I finish brushing my teeth before bed.