Hey, thanks for the report. This is a known client-side issue. The reconnection state gets stuck and doesn’t recover, even though the backend is working fine.
A few things to try:
Update Cursor. You’re on 3.0.9 from April 3, and there have been fixes since then. Help > Check for Updates.
Disable HTTP/2. This has helped other users with similar issues:
Open App Settings Cmd+,
Search for HTTP/2
Turn on Disable HTTP/2
Fully restart Cursor Cmd+Q, then open it again
Network Diagnostics. Cursor Settings > Network > Run Diagnostics. Send the results if you see any errors.
A couple questions:
Are you using a VPN, proxy, or firewall?
Is this a home network or a corporate or university network?
When it happens again, copy the Request ID from the chat menu three dots in the top right > Copy Request ID. The UUID you sent is the conversation ID, but for debugging we need the Request ID.
The team is aware of this class of bugs. Let me know if the steps above help.
From the diagnostics, it looks like Cursor can’t connect to some domains API, Ping, Chat, Agent, while the rest are OK. But you’re on 3.0.16, and since then we’ve shipped a few updates with reconnecting fixes.
First, please update to the latest version: Help > Check for Updates (or download it from Cursor · Download). After updating, run Network Diagnostics again.
If it’s still happening, send a new diagnostics screenshot and the Request ID (three dots in the top-right of the chat > Copy Request ID).
