[BUG] Cowork hangs on "Starting up…" indefinitely + every message crashes — Windows 10 Pro, Desktop 1.8555.2.0
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
Cowork is completely unusable on my machine. Existing sessions hang on "Starting up…" indefinitely (8+ minutes with no progress), and every message I attempt to send (even when starting brand new Cowork sessions) crashes the Cowork session immediately.
Regular Claude chat works fine — only Cowork is affected.
What Should Happen?
Expected Behavior
Cowork should start a working session and accept messages without crashing.
Error Messages/Logs
2026-05-23 22:31:56 [info] [VM:start] Beginning startup
2026-05-23 22:31:58 [info] [VM:steps] vm_boot completed (334ms)
2026-05-23 22:32:06 [info] [VM] Network status: CONNECTED
2026-05-23 22:32:09 [info] [VM] API reachability: REACHABLE
2026-05-23 22:32:11 [info] [VM:steps] sdk_install completed (4027ms)
2026-05-23 22:32:11 [info] [VM:start] Startup complete, total time: 14891msSteps to Reproduce
Steps to Reproduce
Open Claude Desktop 1.8555.2.0 on Windows 10 Pro
Open Cowork
Start a new session with any model (reproduced on Opus 4.7)
Send any message
Observe: session shows "Starting up…" spinner indefinitely; any sent message crashes the session
Claude Model
Opus
Is this a regression?
Yes, this worked in a previous version
Last Working Version
Exact previous version unknown — Claude Desktop auto-updates and the previous build was replaced. Cowork worked reliably for me up until very recently. Earliest log evidence of successful Cowork spawn is on version 1.2278.0.0 (2026-04-13), but Cowork continued working through multiple subsequent updates before breaking.
Claude Code Version
Current broken version is 1.8555.2.0.
Platform
Anthropic API
Operating System
Windows
Terminal/Shell
PowerShell
Additional Information
VM boots cleanly, network connects, API is reachable, SDK 2.1.149 installs successfully. The crash appears to happen at the task spawn layer after the VM is healthy — consistent with the pattern in #49367 and #49485, but on Windows 10 Pro rather than macOS.
I'll attach the full cowork_vm_node.log to this issue.
Possibly relevant: Windows 10 Pro rather than Windows 11. The Cowork documentation primarily references Windows 11, but Windows 10 Pro has the required Hyper-V components and the VM service runs without error, so the failure appears to be downstream of platform compatibility. Using Windows PowerShell 5.1.19041.6456
3 Comments
Found 3 possible duplicate issues:
This issue will be automatically closed as a duplicate in 3 days.
🤖 Generated with Claude Code
This was incorrectly marked as a duplicate of #61489. The issues are not the same — the bot appears to have pattern-matched on the word "Cowork" and called it a day.
#61489 is macOS-specific:
Platform tag: platform:macos
Reporter on macOS 26.3, Apple Silicon, Terminal.app
Symptom: crashes on resuming a long-running session
Root cause: macOS ARG_MAX limit (E2BIG errno) + a 13GB VM bundle accumulated over months
My issue is Windows-specific:
Platform: Windows 10 Pro, Build 19045
Terminal: Windows PowerShell 5.1
Symptom: fresh sessions hang on "Starting up…" indefinitely + every message crashes (not specific to resume)
VM bundle is ~6 weeks old, not 13GB
macOS ARG_MAX/E2BIG does not apply to Windows process spawning
The proposed solutions in #61489 (macOS argument-length serialization, rootfs.img compaction) don't address my failure mode. Please reopen and re-triage.
cc: maintainers — could this be looked at by a human?
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.