[BUG] Claude Cowork conflict with Mac OS 26

[atrankin63-dot]

Preflight Checklist

• I have searched existing issues and this hasn't been reported yet
• This is a single bug report (please file separate reports for different bugs)
• I am using the latest version of Claude Code

What's Wrong?

Cowork workspace fails to load since updating to macOS 26.3. The VM boots successfully but hangs at the guest_vsock_connect step
for 60 seconds, then times out and retries in a loop. This started around 3 Feb after installing macOS 26.3 Beta 3, and persists
on macOS 26.3 final (released 6 Feb).

What Should Happen?

C laude Cowork should launch as usual but it doesn't!

Error Messages/Logs

From cowork_vm_node.log:
  [VM:steps] guest_vsock_connect started
  [VM:start] Connection timeout, last completed step: vm_boot
  [VM:start] Startup failed: Error: VM connection timeout after 60 seconds

  From cowork_vm_swift.log:
  [warn] Timeout waiting for VM to stop (state: 1), proceeding with cleanup

  - Before 3 Feb: Cowork worked normally
  - 3-4 Feb: Intermittent vsock timeouts (macOS 26.3 Beta 3 installed 27 Jan)
  - 6 Feb 13:44: One successful connection (took 8.8s, slower than usual)
  - 6 Feb 15:59+: 10+ consecutive timeouts, no recovery
  - Each failed retry also fails to cleanly stop the previous VM
  - vmnet gateway IP increments each retry (67.1, 68.1, 69.1...) as resources don't release

Steps to Reproduce

Just open Claude Cowork with the latest version of OS26

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

Opus 4.5

Claude Code Version

v2.1.33

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

Log files

• ~/Library/Logs/Claude/cowork_vm_node.log
• ~/Library/Logs/Claude/cowork_vm_swift.log
• ~/Library/Logs/Claude/coworkd.log

[iamkaizenc]

+1

[Cianai]

+2

[rcwhitney]

+3

MacOS 26.2, Claude 1.1.2321 (495628) 2026-02-06T22:01:17.000Z

Everything was working fine last night and then this morning cowork starting having fits. I didn't make any updates to the OS or other files (at least that I'm aware of), nor an update to Claude over night.

I did use claude code to troubleshoot and found a solution to get cowork working again. However, I want to be very clear that I am trusting Claude here and I can't fully confirm that this might not end up with lost data. So use at your own risk, make backups, etc. But here's Claude's advice:

---

Fix: Cowork VM stuck at guest_vsock_connect (macOS 26.x)

If Cowork fails to start with the VM timing out at the guest_vsock_connect step, resetting the VM bundle fixes it:

1. Fully quit Claude Desktop (Cmd+Q, may need twice)

2. Verify it's dead

ps aux | grep -i Claude.app | grep -v grep

3. (Optional) Back up session data if you want to preserve it

mkdir -p ~/Documents/claude-vm-backup
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/sessiondata.img ~/Documents/claude-vm-backup/
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/macAddress ~/Documents/claude-vm-backup/
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/machineIdentifier ~/Documents/claude-vm-backup/

4. Delete the VM bundle

rm -rf ~/Library/Application\ Support/Claude/vm_bundles/

5. Relaunch Claude Desktop - it will re-download the VM image (~10 GB)

First launch may take a few minutes. Cowork conversations are NOT lost (stored separately in local-agent-mode-sessions/).

What you lose: In-VM session state (installed packages, files created within Cowork sessions). What you keep: All Cowork conversation histories, Claude Desktop settings.

[atrankin63-dot]

Thank you. Problem solved!

[bikeracer4487]

Had the same issue, also on MacOS 26.2. Deleting the VM bundle and restarting Claude Desktop did the trick for me as well.

[kandandidat-prog]

Confirming the same issue on macOS 26.3 final (build 25D5112c), MacBook Pro, Claude Desktop latest version.

Logs from cowork_vm_swift.log showed the identical pattern:

• VM boots successfully
• guest_vsock_connect started — then hangs indefinitely
• Timeout after ~4 minutes, retry attempts with gateway IP incrementing (192.168.65.1 → 192.168.66.1)
• Timeout waiting for VM to stop (state: 1), proceeding with cleanup
• No successful connections

VM config: 4 CPUs, 4GB RAM, EFI/GRUB boot, VZVmnetNetworkDeviceAttachment (shared mode).

Fix that worked for me: Resetting the VM bundle resolved the issue.

1. Fully quit Claude Desktop (Cmd+Q)
2. Back up session data: cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/sessiondata.img ~/Documents/claude-vm-backup/
3. Delete VM bundle: rm -rf ~/Library/Application\ Support/Claude/vm_bundles/
4. Relaunch Claude Desktop (re-downloads ~10 GB VM image)

Cowork started normally after this. Conversation histories were preserved.

[qosmonot]

Thanks @rcwhitney, and Claude I guess. had the same issue, I believe the same timeline as the OP, and this fixed it

[fruskate]

same issue MacOS 26.2

[teast5]

Confirmed fix on macOS 26.2 (Tahoe), Claude 1.1.2321 — with full task recovery

Adding to @rcwhitney's and @MaverickKB's findings. I was hitting the same guest_vsock_connect timeout on macOS 26.2 since Saturday night. Rebooting, reinstalling, disabling MCP servers — nothing worked.

The VM bundle delete fix from this thread resolved the boot issue, but the critical missing piece is backing up and restoring sessiondata.img to preserve your Cowork tasks.

Without the restore, all existing tasks throw:

RPC error: ensure user: user dazzling-nifty-thompson should already exist but does not, 
and recovery failed: home directory /sessions/dazzling-nifty-thompson does not exist

Every task becomes orphaned and has to be manually recreated. If you have mature tasks with significant context, that's a painful loss.

Full fix with task recovery:

Step 1 — Quit Claude and back up session data:

osascript -e 'quit app "Claude"'
ps aux | grep -i Claude.app | grep -v grep
mkdir -p ~/Documents/claude-vm-backup
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/sessiondata.img ~/Documents/claude-vm-backup/
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/macAddress ~/Documents/claude-vm-backup/
cp ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/machineIdentifier ~/Documents/claude-vm-backup/

Step 2 — Delete the VM bundle and relaunch:

rm -rf ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle
open -a Claude

Wait for the ~10 GB VM image to re-download and Cowork to become available.

Step 3 — Restore session data:

osascript -e 'quit app "Claude"'
ps aux | grep -i Claude.app | grep -v grep
cp ~/Documents/claude-vm-backup/sessiondata.img ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/sessiondata.img
cp ~/Documents/claude-vm-backup/macAddress ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/macAddress
cp ~/Documents/claude-vm-backup/machineIdentifier ~/Library/Application\ Support/Claude/vm_bundles/claudevm.bundle/machineIdentifier
open -a Claude

All 8 of my existing Cowork tasks came back fully functional.

What this preserves: All task session state, installed packages, working files inside the sandbox.
What gets replaced: Only the rootfs (the Linux OS image), which was the corrupted component.

+1 to @MaverickKB's recommendation for a "Reset Workspace" button in the UI. This shouldn't require Terminal archaeology.

[anotherpanacea-eng]

MacOS 15.7.4 (24G517), same vsock timeout, resolved by deleting vm_bundles