Describe the Bug
When enabling Auto-Run Mode in Cursor (to allow agents to execute commands automatically), the commands are indeed executed without confirmation, which is great. However, after the execution finishes, the agent still gets stuck waiting for user input (e.g. “Move to background” or “Skip”) — as shown in the attached screenshot.
This behavior defeats part of the purpose of Auto-Run, especially for workflows intended to be fully autonomous or hands-free.
Steps to Reproduce
1. Enable Auto-Run Mode in Cursor settings.
2. Trigger a command (e.g. file read, grep, or json write).
3. Observe that while the command runs automatically, the agent pauses after execution awaiting user interaction.
Expected Behavior
After a command is auto-executed in Auto-Run Mode, the agent should proceed automatically without pausing for manual confirmation.
Screenshots / Screen Recordings
Operating System
MacOS
Current Cursor Version (Menu → About Cursor → Copy)
版本: 1.1.6 (Universal)
提交: 1.96.2
日期: 5b19bac7a947f54e4caa3eb7e4c5fbf832389850
Electron: 2025-06-25T02:16:57.571Z
ElectronBuildId: 34.5.1
Chromium: undefined
Node.js: 132.0.6834.210
V8: 20.19.0
OS: 13.2.152.41-electron.0
Does this stop you from using Cursor
Yes - Cursor is unusable