Internal Error An unexpected error occurred on our servers

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

Internal Error
An unexpected error occurred on our servers. Please try again, or contact support if the issue persists.
Copy Request (7c2e1152-9a4a-4544-b53d-cee10f9f9c47)

Steps to Reproduce

The above error always occurs when chatting in the agent window.

Expected Behavior

An error is always occurring with the inquiries I request.
Responses must be made without errors.

Screenshots / Screen Recordings

Operating System

Windows 10/11

Version Information

Version: 3.9.16 (user setup)
VS Code Extension API: 1.105.1
Commit: 042b3c1a4c53f2c3808067f519fbfc67b72cad80
Date: 2026-06-27T06:41:01.941Z
Layout: editor
Build Type: Stable
Release Track: Default
Electron: 40.10.3
Chromium: 144.0.7559.236
Node.js: 24.15.0
V8: 14.4.258.32-electron.0
xterm.js: 6.1.0-beta.256
OS: Windows_NT x64 10.0.26200

For AI issues: which model did you use?

Model name : Composer 2.5 Fast

For AI issues: add Request ID with privacy disabled

Request ID: 7c2e1152-9a4a-4544-b53d-cee10f9f9c47
[internal] Failed to run step, exceeded max retries
RetriableError: [internal] Failed to run step, exceeded max retries
at uAS (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12056:27795)
at EMc (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12056:26530)
at Fny (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12057:11929)
at Object.classify (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12057:16434)
at hAS (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12057:5824)
at async Oqm.runWithSharedTurnRunner (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12057:15786)
at async iH.runAgentLoop (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:24486:3041)
at async Lkh.streamFromAgentBackend (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:24505:13862)
at async Lkh.getAgentStreamResponse (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:24505:22888)
at async IRe.submitChatMaybeAbortCurrent (vscode-file://vscode-app/c:/Users/mrhmr/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:12142:2399)

Does this stop you from using Cursor

Yes - Cursor is unusable

It appears that this error occurs if the <CLAUDE.md> file is large and the process continues within a single Agent window.

It is working now after I reduced the size of the <CLAUDE.md> file, restarted the IDE, and ran it in a new window.

For reference, errors continue to occur in Agent windows where the error has already occurred.

Hey there! Thanks for coming back to share what fixed it, that’s really helpful for anyone else who runs into this.

Your diagnosis is spot on. Instruction files like CLAUDE.md, along with any always-on rules, get sent with every request and count toward the model’s context budget. When one of them grows large and a single agent conversation keeps going, the combined context can push a request past the model’s limit. Once a chat crosses that line, that same thread tends to keep failing on every follow-up, which is exactly why the old window stayed broken while a fresh one worked.

So the two things you did are the right moves:

  • Keep CLAUDE.md (and any rules) lean and focused. Beyond avoiding this error, a tighter instructions file also helps the agent stay on track. A few pointers here: Rules | Cursor Docs
  • Start a new chat for a new task instead of continuing one long thread. An errored conversation won’t recover on its own, so moving to a fresh chat is the fix.

This is also something we’re aware of and actively improving on our side, so it should get more forgiving over time. If it comes back in a fresh chat with a trimmed setup, just drop a new request ID here and I’ll take another look.