Agent hanging at Planning next moves due to Indexing

Hi,

I’ve been vibe coding with Cursor, and until recently, agents have been working well. I don’t know if there’s a bug with the latest upgrade or if the AI servers are just getting busy. It seems that over the last few upgrades, agent access has become inconsistent. Now, more often than not, trying to invoke an agent leaves me at “Planning next moves” indefinitely.

I’m using

Version: 2.3.34 (system setup)
VSCode Version: 1.105.1
Commit: 643ba67cd252e2888e296dd0cf34a0c5d7625b90
Date: 2026-01-10T21:17:10.428Z
Electron: 37.7.0
Chromium: 138.0.7204.251
Node.js: 22.20.0
V8: 13.8.258.32-electron.0
OS: Windows_NT x64 10.0.19045

I’ve tried troubleshooting the Network as a possible issue, but that does not seem to be the issue. However, when I run diagnostics, everything checks okay except DNS, SSL, HTTP/2; they hang. I’ve tried changing to HTTP/1.1 but still SSL and DNS hang. I’ve used curl commands to check Certs, and reachability, and everything seems to check out. I’ve uninstalled anti-virus software to see if that was the issue, but it behaves the same. I’ve looked at the developer logs and don’t see any error messages.

This is an edited addition to this post from earlier today. It seems that this agent stuck in Planning next moves is due to Indexing. As soon as Indexing completes (I happen to catch it while troubleshooting) the agent responded. I was under the impression that indexing would happen in another thread but it seems to block the agent.

I also noticed that I’m only indexing 48 files but restarting Cursor many times throughout the day seemed to show it would start Indexing every time it opened. Shouldn’t indexing only continue if there’s changes?

Please help me solve this.

Steve

Hey, thanks for the detailed report. You identified the issue correctly. It’s actually related to indexing, which is blocking the agent.

This is a known issue and the team is currently investigating it. A few users have reported similar behavior:

About the DNS/SSL/HTTP/2 diagnostics hanging, it could be related to a corporate network or VPN. But if the curl commands work, it’s probably not the main cause.

The team is working on a fix.

Hi Dean,

Thanks for the quick response. I should add that I’m using Cursor to program Raspi Zero 2W via an SSH connection. I love the product and use it quite often. Admittedly, since I’m not a programmer, I rely exclusively on AI, so finding a solution would be great. I thought maybe it was a split-tunnel issue related to the networking, but I came to the same conclusion as you mentioned, and it is related to Indexing and not the misleading network diagnostics. It is strange that this morning, when I came back to it and brought everything up, the Indexing worked and the AI agent as well. I made a bunch of code changes and eventually (and regrettably) had to reboot the Pi. Then the problem returned. This is a trend that I observed. That is, it’s working, then a reboot and reconnection break indexing and the AI agent.

If you’d like me to run more tests to help you troubleshoot, please let me know. I’m happy to help.

Sincerely,
Steve

1 Like

Hi Dean,

I upgraded to version .35 this morning, and both indexing and AI agent access have been performing much better. While the network diagnostics still behave the same way, indexing has been much more consistent.

I have noticed that if I reboot the SSH endpoint, I need to restart Cursor to get indexing to work again.

Thank you for getting me back up and running so quickly.

Best regards,
Steve Schubert

2 Likes

I am still experiencing this same issue.
I was trying to restart the application, and I will restart my whole computer, but it just infinitely sticks in the planning next moves, and nothing is happening.

Version: 2.4.7 (Universal)
VSCode Version: 1.105.1
Commit: ca0f9bf806f235ea014a22712cbcbf5e88ca77e0
Date: 2026-01-20T20:52:38.077Z
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: Darwin arm64 25.1.0

Yeah. I started a new project on an existing system. Again, using a Raspberry Pi & ran into the same problem. I’m starting to believe that I was just getting lucky before. I tried duplicating what was working to no avail. It seemed more often than not. It would get stuck. Maybe twice during the whole day sync worked & AI agent could be used. Notably anytime network diagnostic fully worked so did sync.