After updating Cursor, it is no longer possible to terminate the processes started by Cursor

Support Bug Reports
,
1

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

Before the update, I was able to stop processes launched by Cursor and continue to the next step.

After the update, however, I can no longer stop those processes. They remain hung for more than 10 minutes. The ability to provide input manually or interrupt the process with Ctrl+C has also been removed.

I can stop the entire request/response session, but that is not what I want.

Cursor frequently runs benchmark tests that require user input. When I stop those processes, Cursor understands the issue and continues without requiring user input. However, no matter how many times I report this, Cursor still fails to fix the underlying problem.
Cursor says this, but it never actually fixes the issue and keeps making the same mistake every time:

“RunAll doesn’t pass --filter, so it waits for interactive input. When running everything, it’s recommended to explicitly specify --filter "*" for each project, as shown above.”

Not being able to stop processes that Cursor launches repeatedly is a serious issue. Why was this feature removed? I don’t understand the reasoning behind this decision.

Steps to Reproduce

I asked it to benchmark several features of an open-source project.

Screenshots / Screen Recordings

The [Stop] button is gone!

커서_hang.png
커서_hang.png1162×787 85.8 KB

Operating System

Windows 10/11

Version Information

Version: 3.9.16 (system setup)
VS Code Extension API: 1.105.1
Commit: 042b3c1a4c53f2c3808067f519fbfc67b72cad80
Date: 2026-06-27T06:41:01.941Z
Layout: editor
Build Type: Stable
Release Track: Default
Electron: 40.10.3
Chromium: 144.0.7559.236
Node.js: 24.15.0
V8: 14.4.258.32-electron.0
xterm.js: 6.1.0-beta.256
OS: Windows_NT x64 10.0.26100

For AI issues: which model did you use?

Auto

Does this stop you from using Cursor

No - Cursor works, but with this issue

5

Hey, thanks for the report and the screenshot. I can see the benchmark got stuck waiting for an interactive selection, and in the new terminal card there’s no input field and no Ctrl+C. You only get the “…” menu with Auto-Run options like Allowlist, Auto-review, Run Everything, and Copy Command.

One important detail: the agent shell card was redesigned, so you can’t directly interact with a running command inside chat anymore. There’s a separate Run in background button that can move the command to the bottom terminal pane, where you can type input or stop the process, but it doesn’t show up in all setups. If you don’t have it, there isn’t a clean way to send input to a stuck process right now. This, and the current Stop behavior canceling the whole turn instead of just the command, is on our radar, but I can’t share an ETA for bringing back the old behavior yet.

What fixes your case right now:

  • Pass --filter "*" (or a specific filter) in the launch command so RunAll doesn’t switch to interactive mode and hang. The output also suggests this.
  • Or run interactive commands in a normal terminal outside the agent, where stdin and Ctrl+C work normally.

I know this is a workaround, not the old workflow coming back. Let me know if it helped.