500 Error in Background Composer on Cursor Web

Where does the bug appear (feature/product)?

Background Agent (GitHub, Slack, Web, Linear)

Describe the Bug

Hi Cursor team,

I’m experiencing issues while using the Cursor web app. Whenever I try to work with selected models (e.g., gpt-5 and claude-4.1-opus), I see console errors like:

POST https://cursor.com/api/auth/notifyBackgroundComposerShown 500 (Internal Server Error)
Failed to fetch detailed diff
cannot decode message aiserver.v1.GetOptimizedDiffDetailsResponse from JSON

Additional details:

  • Affected bcId: bc-badc14fd-2c77-46d9-a5a8-50165140b449
  • I’m using only the Cursor web app; no active agents.
  • I’ve tried reloading the page and clearing cache, but the issue persists.
  • Browser: Chrome
  • OS: Windows 10 ( MiniOS10 )
    Thank you in advance for your help.

Steps to Reproduce

Open Cursor web app.

Select one or more models (gpt-5, claude-4.1-opus).

Observe the console logs for errors (500, Failed to fetch detailed diff, etc.).

Attempt to use Background Composer features — the request fails.

Expected Behavior

The web app should successfully fetch diffs and start Background Composer without returning 500 errors. Streaming or polling for updates should work as expected.

Operating System

Windows 10/11

Current Cursor Version (Menu → About Cursor → Copy)

I am not having.

For AI issues: which model did you use?

Any model

For AI issues: add Request ID with privacy disabled

Background Agent: bc-badc14fd-2c77-46d9-a5a8-50165140b449

Additional Information

The error occurs with any selected model.

No active agents are used.

I’ve tried clearing cache and reloading the page; issue persists.

Cursor is unusable in this state.

Does this stop you from using Cursor

Yes - Cursor is unusable

@samubv thank you so much for reporting so detailed. Forwarded the info to the team for checking.

@samubv could you check if this still occurs for you? Please let me know if it does so the team can check it further.

@condor Hello! The same problem still persists.

So, is this happening to just me or someone else?

Checking with the team. Apologies for the delay.

This topic was automatically closed 22 days after the last reply. New replies are no longer allowed.