Claude Desktop Cowork: HCS 0x800707de Construct failure due to duplicate Plan9 shares

Resolved 💬 19 comments Opened Mar 7, 2026 by faisal7272000-cmyk Closed Mar 20, 2026

Bug Description

Claude Desktop's "cowork" workspace fails to start with HCS error 0x800707de during the VM Construct phase. The root cause appears to be duplicate Plan9 shares in the HCS VM configuration.

Error Message

Failed to start Claude's workspace
HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}

Root Cause Evidence

From %APPDATA%\Claude\logs\cowork-service.log, the HCS config contains two identical Plan9 share entries:

"Plan9":{"Shares":[
  {"AccessName":"home","Name":"home","Path":"C:\\Users\\moham","Port":9999},
  {"AccessName":"home","Name":"home","Path":"C:\\Users\\moham","Port":9999}
]}

The HcsCreateComputeSystem call itself succeeds (hr=0x0), but HcsWaitForOperationResult fails (hr=0x800707de) during the Construct phase — consistent with a duplicate resource conflict.

Environment

OS: Windows 11 Home 10.0.26200
Claude Desktop version: Claude_1.1.5368.0_x64 (Windows Store)
VM SDK version: 2.1.63
Windows features enabled: HypervisorPlatform, VirtualMachinePlatform, Windows Subsystem for Linux
Containers feature: Not available (Windows 11 Home)
Hypervisor launch type: Auto

Troubleshooting Performed (all ineffective)

Restarted PC multiple times
Reinstalled workspace via Claude Desktop UI
wsl --shutdown + Restart-Service vmcompute -Force
Deleted %APPDATA%\Claude\claude-code-vm, vm_bundles, and local-agent-mode-sessions — provides temporary fix but error returns on next attempt
Installed latest Windows Update + restart
Verified WSL itself works independently (wsl -d Ubuntu-22.04 echo "hello" succeeds)
Verified vmcompute service is running
Verified 550GB free disk space
No stale named pipes or HCS compute systems lingering

Timeline

Last successful cowork session: 2026-03-02 (confirmed in coworkd.log)
Error first appeared: ~2026-03-03
Previous related incident (Feb 14): Different HCS error 0x80070002 ("host share not mounted") — resolved by deleting vm_bundles. That incident created a vm_bundles_corrupt_20260214_173901 folder.

Relevant Log Files

%APPDATA%\Claude\logs\cowork-service.log — contains full HCS config JSON and error
%APPDATA%\Claude\logs\coworkd.log — shows last successful session activity
%APPDATA%\Claude\logs\main.log

View original on GitHub ↗

19 Comments

github-actions[bot] · 4 months ago

Found 1 possible duplicate issue:

https://github.com/anthropics/claude-code/issues/31691

This issue will be automatically closed as a duplicate in 3 days.

If your issue is a duplicate, please close it and 👍 the existing issue instead
To prevent auto-closure, add a comment or 👎 this comment

🤖 Generated with Claude Code

jackvandy77 · 4 months ago

Same issue — Windows 11 Pro, i9-14900HX, Claude 1.1.5749.0 (Store), Hyper-V fully enabled. Confirmed duplicate Plan9 home shares in cowork-service.log. Logs shared privately via "share your debug logs."

ckirland7317-maker · 4 months ago

Potential Fix - Delete recurring tasks - worked for me

Environment:

Windows 11
Claude Desktop version 1.1.5749.0 (Microsoft Store install)
WSL2 + Ubuntu installed during troubleshooting

Attempted fixes that did NOT work:

Restart-Service vmcompute — failed, service could not be stopped
Start-Service vmcompute — service was already running
wsl --shutdown then relaunch — no effect
Reinstalling workspace via Claude Desktop UI ("reinstall the workspace" link) — no effect
Full reboot — no effect
Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All — already enabled
Enable-WindowsOptionalFeature -Online -FeatureName VirtualMachinePlatform -All — already enabled
Enable-WindowsOptionalFeature -Online -FeatureName HypervisorPlatform -All — already enabled
Enable-WindowsOptionalFeature -Online -FeatureName Containers -All — required reboot, did not fix
Installing WSL2 + Ubuntu from scratch — WSL was missing entirely, installed successfully, but did not fix Cowork
Creating missing WinNAT rule: New-NetNat -Name cowork-vm-nat -InternalIPInterfaceAddressPrefix 172.16.0.0/24 — NAT was missing, created successfully, but did not fix
Renaming sessiondata.vhdx to force reset — no effect
DNS check on vEthernet (cowork-vm-nat) — adapter didn't exist yet, VM wasn't getting that far

Temporary fixes (error returned after next task):

Deleting claude-code-vm and vm_bundles folders — Cowork worked until the next task was started, then error returned
Deleting claude-code-vm and vm_bundles with WinNAT rule in place — same result, temporary only

THE FIX:

Deleted the recurring scheduled Cowork task → rebooted → Cowork launched and ran successfully with no errors

The scheduled task appears to be triggering the VM to spin up in a conflicted state on startup, causing duplicate Plan9 share entries in the HCS VM configuration. Removing the scheduled task and rebooting resolved the issue permanently.

pomomoh · 4 months ago

same issue. removing the scheduled task as @ckirland7317-maker discussed seems to have solved the problem. I've not restarted app/computer however.

bryant-spec · 4 months ago

2026-03-09 14:50:52 [info] [VM] Loading vmClient (TypeScript) module...
2026-03-09 14:50:52 [info] [VM] Module loaded successfully
2026-03-09 14:50:54 [info] [Bundle:status] rootfs.vhdx missing
2026-03-09 14:50:54 [info] [Bundle:status] rootfs.vhdx missing
2026-03-09 14:50:54 [info] [Bundle:status] rootfs.vhdx missing
2026-03-09 14:50:54 [info] [warm] No warm file found for rootfs.vhdx at current version
2026-03-09 14:50:54 [info] [cache] No cached compressed file for rootfs.vhdx
2026-03-09 14:50:54 [info] rootfs.vhdx not found, downloading...
2026-03-09 14:50:54 [info] Downloading rootfs.vhdx...
2026-03-09 14:50:54 [info] [KernelBugMonitor] State reset
2026-03-09 14:50:54 [info] [VM:start] Beginning startup, bundlePath=C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:50:54 [info] [VM:start] Bundle version: fb30784dadb34104626c8cf6d8f90dd47cd393cc
2026-03-09 14:50:54 [info] [VM:start] VM instance ID: 56e069a4-723a-4dcf-b6af-8699533d85eb
2026-03-09 14:50:54 [info] [VM:steps] download_and_sdk_prepare started
2026-03-09 14:50:54 [info] [downloadVM] Download already in progress, waiting...
2026-03-09 14:50:54 [info] [KernelBugMonitor] State reset
2026-03-09 14:50:54 [info] [VM:start] Beginning startup, bundlePath=C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:50:54 [info] [VM:start] Bundle version: fb30784dadb34104626c8cf6d8f90dd47cd393cc
2026-03-09 14:50:54 [info] [VM:start] VM instance ID: 56e069a4-723a-4dcf-b6af-8699533d85eb
2026-03-09 14:50:54 [info] [VM:steps] download_and_sdk_prepare started
2026-03-09 14:50:54 [info] [downloadVM] Download already in progress, waiting...
2026-03-09 14:51:06 [info] [startVM] VM startup already in progress, waiting...
2026-03-09 14:51:54 [info] rootfs.vhdx.zst checksum validated
2026-03-09 14:51:54 [info] [downloadVM] Saved compressed rootfs.vhdx to cache
2026-03-09 14:51:54 [info] [Bundle:status] vmlinuz missing
2026-03-09 14:51:54 [info] [warm] No warm file found for vmlinuz at current version
2026-03-09 14:51:54 [info] [cache] No cached compressed file for vmlinuz
2026-03-09 14:51:54 [info] vmlinuz not found, downloading...
2026-03-09 14:51:54 [info] Downloading vmlinuz...
2026-03-09 14:51:54 [info] vmlinuz.zst checksum validated
2026-03-09 14:51:54 [info] [downloadVM] Saved compressed vmlinuz to cache
2026-03-09 14:51:54 [info] [Bundle:status] initrd missing
2026-03-09 14:51:54 [info] [warm] No warm file found for initrd at current version
2026-03-09 14:51:54 [info] [cache] No cached compressed file for initrd
2026-03-09 14:51:54 [info] initrd not found, downloading...
2026-03-09 14:51:54 [info] Downloading initrd...
2026-03-09 14:51:58 [info] initrd.zst checksum validated
2026-03-09 14:51:58 [info] [downloadVM] Saved compressed initrd to cache
2026-03-09 14:51:58 [info] All files ready in C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:51:58 [info] [VM:steps] download_and_sdk_prepare completed (63879ms)
2026-03-09 14:51:58 [info] [VM:steps] load_swift_api started
2026-03-09 14:51:58 [info] [VM:steps] download_and_sdk_prepare completed (63791ms)
2026-03-09 14:51:58 [info] [VM:steps] load_swift_api started
2026-03-09 14:51:58 [info] [VM:steps] load_swift_api completed (0ms)
2026-03-09 14:51:58 [info] [Callbacks] Initializing VM event callbacks...
2026-03-09 14:51:58 [info] [VM:steps] load_swift_api completed (1ms)
2026-03-09 14:51:58 [info] [Callbacks] VM event callbacks initialized
2026-03-09 14:51:58 [info] [VM:start] Copying smol-bin.x64.vhdx to bundle: C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\smol-bin.x64.vhdx -> C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle\smol-bin.vhdx
2026-03-09 14:51:58 [info] [VM:start] Copying smol-bin.x64.vhdx to bundle: C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\smol-bin.x64.vhdx -> C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle\smol-bin.vhdx
2026-03-09 14:51:58 [info] [VM:start] smol-bin.x64.vhdx copied successfully
2026-03-09 14:51:58 [info] [VM:start] Configuring Windows VM service...
2026-03-09 14:51:58 [info] [VM:start] smol-bin.x64.vhdx copied successfully
2026-03-09 14:51:58 [info] [VM:start] Configuring Windows VM service...
2026-03-09 14:51:58 [info] [VM:start] Windows VM service configured
2026-03-09 14:51:58 [info] [VM:start] Windows VM service configured
2026-03-09 14:51:59 [error] [VM:start] VM boot failed: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:51:59 [info] Dispatching startup error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:51:59 [error] [VM:start] Startup failed: Error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
at kLt (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:914:3941)
at async Object.startVM (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1358:25816)
at async C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:7:255400
at async Session.<anonymous> (node:electron/js2c/browser_init:2:116443) {
[stack]: [Getter/Setter],
[message]: 'HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}'
}
2026-03-09 14:51:59 [info] [VM:start] Auto-reinstalling workspace after startup failure
2026-03-09 14:51:59 [info] [deleteVMBundle] Deleting reinstall files from C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:51:59 [info] [deleteVMBundle] Reinstall files deleted (sessiondata.img and rootfs.img.zst preserved)
2026-03-09 14:51:59 [info] [VM:shutdown] App quit, stopping VM (instance: 56e069a4-723a-4dcf-b6af-8699533d85eb)...
2026-03-09 14:51:59 [error] [VM:start] VM boot failed: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:51:59 [info] Dispatching startup error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:51:59 [error] [VM:start] Startup failed: Error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
at kLt (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:914:3941) {
[stack]: [Getter/Setter],
[message]: 'HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}'
}
2026-03-09 14:51:59 [info] [VM:start] Skipping auto-reinstall (already attempted once)
2026-03-09 14:51:59 [info] [VM:shutdown] Completed in 153ms
2026-03-09 14:52:02 [info] [VM] Loading vmClient (TypeScript) module...
2026-03-09 14:52:02 [info] [VM] Module loaded successfully
2026-03-09 14:52:04 [info] [KernelBugMonitor] State reset
2026-03-09 14:52:04 [info] [VM:start] Beginning startup, bundlePath=C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:52:04 [info] [VM:start] Bundle version: fb30784dadb34104626c8cf6d8f90dd47cd393cc
2026-03-09 14:52:04 [info] [VM:start] VM instance ID: 292f2d29-9a91-49f9-94bc-66d04832e417
2026-03-09 14:52:04 [info] [VM:steps] download_and_sdk_prepare started
2026-03-09 14:52:04 [info] [VM:steps] download_and_sdk_prepare completed (4ms)
2026-03-09 14:52:04 [info] [VM:steps] load_swift_api started
2026-03-09 14:52:04 [info] [VM:steps] load_swift_api completed (0ms)
2026-03-09 14:52:04 [info] [Callbacks] Initializing VM event callbacks...
2026-03-09 14:52:04 [info] [Callbacks] VM event callbacks initialized
2026-03-09 14:52:04 [info] [VM:start] Copying smol-bin.x64.vhdx to bundle: C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\smol-bin.x64.vhdx -> C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle\smol-bin.vhdx
2026-03-09 14:52:04 [info] [KernelBugMonitor] State reset
2026-03-09 14:52:04 [info] [VM:start] Beginning startup, bundlePath=C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle
2026-03-09 14:52:04 [info] [VM:start] Bundle version: fb30784dadb34104626c8cf6d8f90dd47cd393cc
2026-03-09 14:52:04 [info] [VM:start] VM instance ID: 292f2d29-9a91-49f9-94bc-66d04832e417
2026-03-09 14:52:04 [info] [VM:steps] download_and_sdk_prepare started
2026-03-09 14:52:04 [info] [VM:steps] download_and_sdk_prepare completed (3ms)
2026-03-09 14:52:04 [info] [VM:steps] load_swift_api started
2026-03-09 14:52:04 [info] [VM:steps] load_swift_api completed (0ms)
2026-03-09 14:52:04 [info] [VM:start] Copying smol-bin.x64.vhdx to bundle: C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\smol-bin.x64.vhdx -> C:\Users\bryan\AppData\Roaming\Claude\vm_bundles\claudevm.bundle\smol-bin.vhdx
2026-03-09 14:52:04 [info] [VM:start] smol-bin.x64.vhdx copied successfully
2026-03-09 14:52:04 [info] [VM:start] Configuring Windows VM service...
2026-03-09 14:52:04 [info] [VM:start] smol-bin.x64.vhdx copied successfully
2026-03-09 14:52:04 [info] [VM:start] Configuring Windows VM service...
2026-03-09 14:52:05 [info] [VM:start] Windows VM service configured
2026-03-09 14:52:05 [info] [VM:start] Windows VM service configured
2026-03-09 14:52:05 [error] [VM:start] VM boot failed: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:52:05 [info] Dispatching startup error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:52:05 [error] [VM:start] Startup failed: Error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
at kLt (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:914:3941)
at async Object.startVM (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1358:25816)
at async C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:7:255400
at async Session.<anonymous> (node:electron/js2c/browser_init:2:116443) {
[stack]: [Getter/Setter],
[message]: 'HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}'
}
2026-03-09 14:52:05 [info] [VM:start] Skipping auto-reinstall (already attempted once)
2026-03-09 14:52:05 [error] [VM:start] VM boot failed: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:52:05 [info] Dispatching startup error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
2026-03-09 14:52:05 [error] [VM:start] Startup failed: Error: HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}
at kLt (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:914:3941)
at async Object.Y6t [as ensureReady] (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1264:670)
at async pFe.prepare (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1261:18710)
at async C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1261:18916
at async pFe.withMutex (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1261:31856)
at async rUt (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1264:3727)
at async Object.addMarketplace (C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:1264:11862)
at async C:\Program Files\WindowsApps\Claude_1.1.5749.0_x64__pzs8sxrjxfjjc\app\resources\app.asar\.vite\build\index.js:7:277482
at async Session.<anonymous> (node:electron/js2c/browser_init:2:116443) {
[stack]: [Getter/Setter],
[message]: 'HCS operation failed: failed to create compute system: HcsWaitForOperationResult failed with HRESULT 0x800707de: {"Error":-2147022882,"ErrorMessage":"","Attribution":[{"OperationFailure":{"Detail":"Construct"}}]}'
}

MidnightOnMars · 4 months ago

Additional data point: logs capture the exact transition from working to broken in the same session.

Claude Desktop v1.1.5368, Windows 11. Cowork was functioning normally until today.

Working state (13:00:30 and 13:00:36 local time):
Single Plan9 share in the HCS config JSON:

"Plan9":{"Shares":[{"AccessName":"home","Name":"home","Path":"C:\\Users\\<user>","Port":9999}]}

HcsWaitForOperationResult returns hr=0x0 both times. VM starts successfully.

Broken state (15:01:00 local time):
The configure RPC fires twice within ~1ms. The cowork-service.log shows the race condition clearly:

2026/03/09 15:01:00.144919 [VM] Plan9 share added: home -> C:\Users\<user> (port=9999, readOnly=false)
2026/03/09 15:01:00.145452 [Server] Configuration complete for <user>:C:\Users\<user>\AppData\Roaming\Claude
2026/03/09 15:01:00.145452 [Server] Sent response: success=true
2026/03/09 15:01:00.145984 [VM] Plan9 share added: home -> C:\Users\<user> (port=9999, readOnly=false)
2026/03/09 15:01:00.145984 [Server] Configuration complete for <user>:C:\Users\<user>\AppData\Roaming\Claude
2026/03/09 15:01:00.145984 [Server] Sent response: success=true

From this point forward, the HCS config JSON contains the duplicated share:

"Plan9":{"Shares":[{"AccessName":"home","Name":"home","Path":"C:\\Users\\<user>","Port":9999},{"AccessName":"home","Name":"home","Path":"C:\\Users\\<user>","Port":9999}]}

Every subsequent HcsWaitForOperationResult returns hr=0x800707de with {"OperationFailure":{"Detail":"Construct"}}. The service retries repeatedly (15:01:00, 15:01:13, 15:01:23, 15:01:25, 15:01:27, 15:01:31) and fails identically each time.

Context: This occurred after a recurring Cowork task executed. The recurring task was deleted and the app was restarted, but the duplicate configure call persists on every relaunch. Deleting claude-code-vm, vm_bundles, and local-agent-mode-sessions did not resolve it -- the app regenerates the double-configure on fresh state.

Root cause: Race condition in the Electron app's configure RPC to cowork-svc. The app sends two configure messages in rapid succession. The service appends both Plan9 shares without deduplication, poisoning the HCS config for the lifetime of the process.

Suggested fix (either would resolve this):

Make the service's configure handler idempotent: Deduplicate Plan9 shares by Name before appending to the shares array. This is the defensive fix and would prevent this class of bug regardless of how many times configure is called.
Fix the Electron app to not double-send: Add a guard (debounce, request-in-flight flag, or similar) so the configure RPC cannot fire twice for the same session. This addresses the root cause.

Both are cheap and not mutually exclusive. The idempotent handler is the safer belt-and-suspenders approach.

imsopov · 4 months ago

I've been getting the same issue here. PC was left on over the weekend and was working fine this morning, accidently kicked the power plug and had to boot up the PC again, Claude Cowork hasn't worked since booting.

Uninstalling and reinstalling has fixed it for now.

Stu-10038 · 4 months ago

Hi, I had cowork as unusable as it would not load the VM.
After much research I found that running these commands in admin powershell fixed it.
In case you also have this frustration run this is admin powershell on windows.

Workaround that fixed it for me: restart the CoworkVMService from PowerShell (run the 2 lines below as Administrator):

Stop-Service CoworkVMService -Force
Start-Service CoworkVMService

Then relaunch Claude Desktop.

ScienceDuck · 4 months ago

Hi, I had cowork as unusable as it would not load the VM. After much research I found that running these commands in admin powershell fixed it. In case you also have this frustration run this is admin powershell on windows. Workaround that fixed it for me: restart the CoworkVMService from PowerShell (run the 2 lines below as Administrator): Stop-Service CoworkVMService -Force Start-Service CoworkVMService Then relaunch Claude Desktop.

Was having the same issue on version 1.1.5749 and this fixed it.

stephenpmills · 4 months ago

hitting it on version 1.1.5749 with the duplicate share visible in the log.

salmonberrynovamvps · 4 months ago

Still occurring after full folder deletion. Confirmed on Windows 11 Pro, AMD Ryzen 7 5800XT, 32GB RAM, Claude Desktop v1.1.5368.
Deleted claude-code-vm, vm_bundles, and local-agent-mode-sessions completely. Claude reinstalled the bundle from scratch. Duplicate Plan9 entry is regenerated fresh every single attempt — this is not a cached config issue, it's being generated by the app itself.
Log confirms identical duplicate on every boot:
"Plan9":{"Shares":[
{"AccessName":"home","Name":"home","Path":"C:\\Users\\leann","Port":9999},
{"AccessName":"home","Name":"home","Path":"C:\\Users\\leann","Port":9999}
]}
HcsCreateComputeSystem succeeds (hr=0x0) but HcsWaitForOperationResult fails (hr=0x800707de) at Construct phase every time. Reinstalling workspace via UI does not resolve. This system had Cowork working previously. Submitted multiple in-app bug reports as well. Waiting on a fix.

Imfineandrew · 4 months ago

experiencing this issue on Windows 10 Home (Build 26200.7840),
Claude Desktop v1.1.6041.0 after reinstalling the app.

hcsdiag list and Get-HnsNetwork both return empty.
WSL2 is installed and working correctly (v2.6.3.0).

Please prioritize a fix as this makes the Cowork feature completely
unusable for paying subscribers.

Imfineandrew · 4 months ago

Temporary workaround that worked for me:

Run the following in PowerShell (as Administrator):

Stop Claude and CoworkVMService:

Stop-Process -Name "Claude" -Force -ErrorAction SilentlyContinue
Stop-Service CoworkVMService -Force -ErrorAction SilentlyContinue

Delete vm_bundles:

Remove-Item -Recurse -Force "$env:APPDATA\Claude\vm_bundles" -ErrorAction SilentlyContinue

Restart vmcompute:

Restart-Service vmcompute -Force

Start CoworkVMService manually:

Start-Service CoworkVMService

Relaunch Claude Desktop

This got Cowork working again for me.
Note: This appears to be a temporary fix only — the error may return on next attempt.

Environment: Windows 10 Home, Build 26200.7840, Claude Desktop v1.1.6041.0

JosephSemet · 4 months ago

Temporary workaround that worked for me: Run the following in PowerShell (as Administrator): 1. Stop Claude and CoworkVMService: Stop-Process -Name "Claude" -Force -ErrorAction SilentlyContinue Stop-Service CoworkVMService -Force -ErrorAction SilentlyContinue 2. Delete vm_bundles: Remove-Item -Recurse -Force "$env:APPDATA\Claude\vm_bundles" -ErrorAction SilentlyContinue 3. Restart vmcompute: Restart-Service vmcompute -Force 4. Start CoworkVMService manually: Start-Service CoworkVMService 5. Relaunch Claude Desktop This got Cowork working again for me. Note: This appears to be a temporary fix only — the error may return on next attempt. Environment: Windows 10 Home, Build 26200.7840, Claude Desktop v1.1.6041.0

This works, thank you.

Disdainz · 4 months ago

Confirmed duplicate Plan9 shares on Windows 10 Home, build 26200, Claude v1.1.5749.0 (Store install)
Same root cause — HCS config contained two identical Plan9 share entries. HcsCreateComputeSystem succeeds (hr=0x0) but HcsWaitForOperationResult fails with 0x800707de at Construct.
Appeared after auto-update from v1.1.5368 to v1.1.5749. Resolved by stopping CoworkVMService, removing vm_bundles, claude-code-vm, and local-agent-mode-sessions from %APPDATA%\Claude, then restarting vmcompute and relaunching Claude Desktop. Bundle rebuilt cleanly with a single Plan9 share entry.

bjc144-svg · 4 months ago

Confirmed duplicate Plan9 shares on my machine as well.

Claude Desktop v1.1.6041.0
Windows 11, Dell Inspiron 15, 13th Gen Intel Core i7-1355U
AzureAD-joined, Intune-managed device

From cowork-service.log:
"Plan9":{"Shares":[
{"AccessName":"home","Name":"home","Path":"C:\\Users\\xxxxxx","Port":9999},
{"AccessName":"home","Name":"home","Path":"C:\\Users\\xxxxxx","Port":9999}
]}

The duplicate share causes HcsWaitForOperationResult to return hr=0x800707de during Construct. Error recurs on every reboot. Temporary workaround of deleting vm_bundles and claude-code-vm resolves it until the next reboot, at which point the duplicate shares reappear in the generated config.

jackvandy77 · 4 months ago

The issued has seemed to resolve for me. My Claude Cowork VM environment has been operational and stable for 6+ hours in a row now.

Imfineandrew · 4 months ago

Resolved.

Thank you ALL.

github-actions[bot] · 3 months ago

This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.