Networking issues on Windows
Issues the maintainers labelled both platform:windows and area:networking.
129 issues · 37 open · 92 resolved (71%) · first seen Feb 26, 2026
Is this getting better or worse?
This class of problem is still growing. 59 new reports in the last 90 days vs 45 in the 90 before — +31%. The open backlog peaked at 37 in 2026-10 and sits at 37 today.
Derived by counting each matching issue from its creation date until it closed. First-party
computation over the anthropics/claude-code tracker — the trend is not published
anywhere upstream.
Which builds report this
90 of these reports name the Claude Code build they were running, spanning 65 releases. Heaviest reporters:
- Claude Code v2.1.233
- Claude Code v2.1.150
- Claude Code v2.1.215
- Claude Code v2.1.227
- Claude Code v2.1.119
- Claude Code v2.1.81
- Claude Code v2.1.266
- Claude Code v2.1.246
Report volume, not a defect count: an issue can mention a build without that build having caused the failure. Versions come from the version string in the bug-report template, matched against real published tags.
Do these get fixed?
Median time to close is 33 days across 92 closures. Of the 92 closures with a recorded reason, 15% were closed as completed and 78 as not-planned or duplicate. The most recent completed fix landed in 2026-10.
Workarounds reported by the community
31 of these issues post a workaround someone says worked. The highest-engagement ones:
- [Feature] Cowork desktop app: support host proxy settings (HTTP/HTTPS/NTLM) for enterprise network environments
- I published a related unofficial Windows Desktop workaround that may help one adjacent case, but I want to be careful about scope: Latest release ZIP: This is not a full fix for Cowork VM egress, NTLM proxy support, or e…
Found in the thread of #45994 · still open - [BUG] Allow network egress - All domains' setting not reflected in session proxy JWT
- …o this is not a tool-layer issue and there is no in-sandbox workaround. For completeness, the form: The denial comes from an upstream gateway : Direct CONNECT against the local proxy: So the local proxy relays and the ga…
Found in the thread of #34690 · still open - [BUG] Cowork Windows: API unreachable from VM behind corporate NTLM proxy (cntlm)
- …h the host's network stack (which already has proxy access) Workaround Using Claude Code CLI directly on Windows, which runs natively and uses the system proxy without issues. Related Issues - 26258 — Similar API unreach…
Found in the description of #29367 · still open - [FEATURE] Remote Control: failure-resistant session model — survive hibernation, auto-reattach, mobile-side recovery
- …The reconnect path exists in embryo; it just doesn't work. Workaround we settled on (mentioned in case it helps others running server workers): we no longer restart the worker on every patch release — the new CLI is dow…
Found in the thread of #78784 · still open - Voice mode fails with 'Voice connection failed' — Cloudflare challenge blocks WebSocket to claude.ai
- …oudflare challenge, similar to how handles CLI connections. Workaround Using a VPN changes the IP and bypasses the Cloudflare challenge, making voice mode work. This confirms the issue is IP/region-based Cloudflare chall…
Found in the description of #34414 · resolved
Excerpts are quoted from the linked issue so you can judge relevance before opening it. Threads that only ask for a workaround, or report that there is none, are excluded. These are community suggestions, not official guidance from Anthropic — read the full thread before applying one.
Matching issues 1–100
Ranked by community engagement (reactions weighted over comments).
- [Feature] Cowork desktop app: support host proxy settings (HTTP/HTTPS/NTLM) for enterprise network environments
- [BUG] Allow network egress - All domains' setting not reflected in session proxy JWT
- [BUG] Cowork Windows: API unreachable from VM behind corporate NTLM proxy (cntlm)
- [BUG] Cowork cloud sandbox: egress allowlist collapses to ~5 hosts despite account setting "All domains"
- [BUG] Claude Code sustained ~80% ECONNRESET failure rate — March 30, 2026 specifically from Boston Area
- [BUG] Persistent ECONNRESET / "Connection lost mid-response" on all Code surfaces while raw HTTPS to api.anthropic.com is healthy (Windows 11, v2.1.229)
- [BUG] Claude Desktop SSH: "Timed out while waiting for handshake" - Windows to Linux, server auth succeeds, getSSHGitInfo fails
- [BUG] I am always getting the API Error: Connection dropped...
- [BUG] API Error: Unable to connect to API (ECONNRESET)
- [FEATURE] Remote Control: failure-resistant session model — survive hibernation, auto-reattach, mobile-side recovery
- [BUG] Every command retries up to 10 times ("Retrying in Xs · attempt X/10") and never gets a response. Happens on every single input.
- [BUG] Claude Code stream dies after first chunk — constant silent stalls, extensive diagnostics attached
- [BUG] SSH remote: message exceeding server line buffer wedges connection in infinite reconnect loop (token too long + stdin replay poison pill)
- Voice mode fails with 'Voice connection failed' — Cloudflare challenge blocks WebSocket to claude.ai
- Remote control reconnection fails permanently after iOS connection drop
- [BUG] Cowork (Windows): device VM egress ignores the account Domain allowlist - 403 blocked-by-allowlist while set to All domains
- Native update/install fails with 'socket connection was closed unexpectedly' behind proxy — undici TLS incompatibility
- [BUG] [Cowork] VM API unreachable on corporate network — no proxy support in VM environment
- [BUG] Remote Control sessions require PC-side input to reconnect after restart; phone-side reconnection fails
- [Bug] Recurring API timeout during "Advising using Opus 4.8" step on Windows with retry loops
- [Bug] Remote control session disconnects after ~5 minutes idle timeout
- [BUG] Remote Control returns HTTP 403 on the in-session `/bridge` path, while `claude remote-control` works on the same machine and account
- [BUG] Claude Code 2.1.228 repeatedly fails with ECONNRESET on Windows 11, while 2.1.220 works normally
- [BUG] [BUG] Persistent ECONNRESET on v2.1.228 (post-fix) — all networks, all local causes eliminated incl. full AV uninstall; claude.ai chat works fine on same account
- Claude Desktop (Cowork): TLS verification failure on bridge.claudeusercontent.com — bundled Node CA bundle issue on Windows 11 [regression ~1 week]
- [BUG] TLS certificate verification fails on Windows unless connected to VPN
- Remote Control fails to connect with "connect timed out" — reproducible, session confirmed alive
- Recurring fetch() socket disconnects in Node SDK — n=81 corpus, cross-client cascade correlated with status.claude.com "Elevated errors" entry
- "API Error: The response stopped arriving" — 33 times in one day, zero on any prior day; error provides no attribution to diagnose it
- Remote Control shows sessions as "Disconnected" while the local claude.exe process is still running
- [BUG] HTTP_PROXY/HTTPS_PROXY stripped from PowerShell/Bash tool subprocess env on Windows (NO_PROXY survives)
- Remote Control silently fails when HTTPS-scanning AV (e.g. Norton 360) intercepts TLS - Node rejects MITM cert chain
- [BUG] API ConnectionRefused and Image Recognition broken after upgrading to 2.1.122 with MiniMax
- Bug: ECONNRESET ('terminated') during streaming with partial data kills the session — no retry, no fallback
- [Bug] No response after several minutes of thinking
- [BUG] Claude Code VSCode extension: ConnectionRefused behind corporate PAC proxy (Windows)
- [BUG] Cowork Windows: API unreachable behind NTLM proxy — no egress proxy implementation on Windows (systematic diagnosis)
- Remote Control disconnects after a few seconds (regression ~March 7-8)
- [BUG] Claude Projects: every cloud thread is server-side bound to the desktop device (config:claude-code-device-bind), which blocks egress writes and Google OAuth refresh; only fix ("Require trusted devices") disables project Remote Control
- [FEATURE] Claude Desktop (MSIX): Support interactive proxy authentication (NTLM) for corporate VPN environments
- [BUG] Cowork cloud sessions ignore "Allow network egress: All domains" and enforce a narrow built-in allowlist; all third-party API calls from scripts fail (403 at CONNECT) on both the cloud shell and the local VM shell
- [BUG] Remote Control SSE channel (worker/events/stream) ignores NODE_EXTRA_CA_CERTS while the API channel honors it — same process, v2.1.201 (it's Node, not Bun)
- [BUG] Title: /remote-control skill immediately disconnects on Windows 11 — "connecting → disconnected" loop
- [BUG] ECONNRESET / "Unable to connect to API" — Claude Code client fails while curl, Node, and the network all succeed (Windows native v2.1.150 + desktop app)
- [BUG] Checking connectivity... hangs indefinitely when using Outline VPN (TUN mode) on Windows 11
- [BUG] git push fails with HTTP 503 on git-receive-pack endpoint via the Claude Code proxy
- [Issue Workaround] Cowork Windows: API unreachable if the device is behind proxy - TUN adapter solution
- [BUG] Claude Code MCP does not use the NO_PROXY setting in environment variables
- [BUG][Windows/Desktop] Code hangs ~7.5 min then ECONNREFUSED; bundled claude.exe works standalone
- [BUG] SSL certificate verification failed when routing via custom inference gateway / corporate proxy
- [BUG] /remote-control times out when re-attaching a session that was previously attached
- [BUG] API Error: "Connection lost mid-response" — server_error on tool-call emission
- [BUG] Consantly getting: API Error: Stream idle timeout - no chunks received
- [BUG] Hung tool calls should be cancellable — a timeout or manual cancel button per tool call
- [FEATURE] Claude Code on Windows silently fails to start - no error message when http_proxy cannot be parsed as URL
- [Windows] Persistent ECONNRESET on streaming API connections due to missing SO_KEEPALIVE (~346s pattern)
- Remote folder connection fails with "Host denied (verification failed)" for specific hostname
- Remote Control: input silently dropped on all clients — messages appear sent but never arrive at local process (regression ~May 25)
- Intermittent `socket connection was closed unexpectedly` (ECONNRESET) on startup — ~40-70% failure rate across parallel startup fetches
- [BUG] Claude Code: messaggi assistant non visualizzati dopo riavvio remote control con VPN spenta
- [BUG] gRPC OTLP exporter ignores otelHeadersHelper-generated headers (request metadata not injected)
- [BUG] Claude code will not connect to antrhropic behind a corporate firewall
- [BUG] Cowork scheduled task: domain api.asaas.com blocked by egress proxy (was working previously)
- Requests silently stall for 900s with no error or retry logged
- Remote Control: sessions attached to a bridge are unrecoverable after the bridge process dies, even when the replacement bridge re-adopts the same environment
- [Bug] ECONNRESET on interactive turns after 15–20 seconds with Claude Sonnet model
- [BUG] Desktop app: long streaming responses drop with ECONNRESET since v1.24012.11 auto-update
- [BUG]: Persistent intermittent ECONNRESET immediately after first streamed chunk, Windows, multiple CLI versions and networks.
- [BUG] Backgrounded/reconnected sessions do not use the configured proxy and fail to connect (ConnectionRefused)
- [BUG] ConnectionRefused on /v1/messages streaming while simple API requests succeed (Windows, Bun)
- [BUG] [BUG] Remote session stuck "Reconnecting" ~95% of the time despite healthy network/server
- [BUG] Cannot use claude code in phpstorm : API error
- Remote control (claude.ai/code) - messages sent from web/phone not reaching CLI session
- [BUG] Socket/connection leak in Windows desktop app — established TCP connections to API grow unbounded over a session
- [BUG] remote-control slow, unresponsive
- [BUG] connectio problem
- [BUG] Remote Control auto-connect (toggle) dies after an overnight idle; only a manual /remote-control re-arms the shared bridge
- [BUG] Remote - from PC (web) to desktop (windows) - only working one way
- Title: Remote control: cloud→local input channel broken (account-scoped);outputchannelworks
- [BUG] Claude Code 2.1.152 returns ConnectionRefused with Anthropic-compatible endpoint while PowerShell and Node HTTPS succeed
- [BUG] Unable to connect to API (ConnectionRefused)
- [Bug] Remote-control notifications not reaching iPhone and responses not syncing to CLI on Windows
- [Bug] Excessive API usage with no response completion
- [Bug] Remote control session disconnects with "reconnecting" message on Windows 10
- [BUG] [BUG] Cowork - objects.anthropic.com returns NXDOMAIN from Israel (DNS resolution fails globally)
- [Bug] Remote control connection repeatedly fails
- Cowork VM networking conflicts with gvisor-tap-vsock (CRC/Podman) on vsock port 1024
- [Bug] SSH/Telnet connections dropping unexpectedly
- [BUG] Title: Sessions stuck on "Can't reach your computer" after reconnects; repeated forced logouts (Windows desktop app)
- [BUG] API Error: Connection refused — a firewall or proxy may be blocking it (ECONNREFUSED)
- [BUG] Cowork/Claude Code egress proxy blocks dev.azure.com with 403 despite domain present in Capabilities network allowlist
- Desktop app: repeated "Request failed · Retrying (n/10)", sometimes needs a full restart
- [BUG] Cowork/cloud sessions: artifact reads stay blocked after adding *.frame.claudeusercontent.com to the environment allowlist
- [Bug] Non-streaming retry receives event-stream response instead of JSON (HTTP 200)
- Subagent stream dies silently on transient network failure — no retry/reconnect, watchdog kills after 600s of silence
- [BUG] Desktop app: long streaming responses drop with ECONNRESET since v1.24012.11 auto-update
- [BUG] claude-sonnet-5[1m] temporarily unavailable (timed out)
- Frequent ECONNRESET disconnects on Windows 11
- [BUG] Remote Control: messages sent from mobile don't reach local session on Windows + WiFi + Cloudflare WARP (MASQUE), no error shown
- [BUG] Connection lost mid-response since Aug 9-10, persists after update to 2.1.227 and model switch