Long-running agent loop prompt appears after update, but I can’t find any way to enable it

Support Bug Reports
1

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

After updating to the latest Cursor version, I saw an in-app message saying I can try “Long-running agent loop”, but I can’t find any UI entry, toggle, or command to enable it.

Steps to Reproduce

Update Cursor to the latest version.

Launch Cursor.

Observe the in-app notification suggesting I can try “Long-running agent loop”.

Look through the Agent UI / settings / command palette for a way to enable it.

Expected Behavior

If the feature is available for my version/account, there should be a clear entry point in the UI (or a documented command) to enable and use it. If it’s not available to me, the prompt shouldn’t appear, or it should explain the requirements.

Screenshots / Screen Recordings

BAG.png
BAG.png262×408 9.33 KB

Operating System

Windows 10/11

Version Information

Version: 2.4.31 (system setup)
VSCode Version: 1.105.1
Commit: 3578107fdf149b00059ddad37048220e41681000
Date: 2026-02-08T07:42:24.999Z
Build Type: Stable
Release Track: Default
Electron: 39.2.7
Chromium: 142.0.7444.235
Node.js: 22.21.1
V8: 14.2.231.21-electron.0
OS: Windows_NT x64 10.0.26200

For AI issues: which model did you use?

any model same

Does this stop you from using Cursor

No - Cursor works, but with this issue

3

Hey, this is now officially available. Long-running agents are a Cloud Agent feature. They run in a remote environment via the web app, not directly in the desktop IDE.

To try it:

It’s available for Ultra users, so you should have access.

More details are in the announcement Introducing Long-running Agents and the blog post https://cursor.com/blog/long-running-agents. You can also check the Cloud Agents docs for setup instructions: Cloud Agents | Cursor Docs

The in-app notice you saw was a bit early. The feature officially launched on February 12. Let us know if you run into any access issues.

4

Very exiting! I just launch my first long running agent to try.

But immediately I do realize that github and cloud agent is bad idea. It works of course, but it has NO MEANS TO TEST the thing before I bring it to local computer with required software. Meaning that the useability as “long running” is very limited.