[BUG] Claude Code segfaults on startup when Anthropic API is unreachable (should fail gracefully)
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?
Claude Code (v2.1.78 and v2.1.76) crashes with Segmentation fault (core dumped) on startup — including claude --version — when the Anthropic API is experiencing issues. Proven by running sudo unshare -n sudo -u $USER ~/.local/share/claude/versions/2.1.76 --version which works perfectly, while the same command with network access segfaults. The binary should display a connection error, not crash.
What Should Happen?
A human-readable error message like "Unable to connect to Anthropic API. Check your connection or visit status.claude.com" and a non-zero exit code. claude --version should never require network access at all.
Error Messages/Logs
$ claude --version
Segmentation fault (core dumped)
# From dmesg:
2.1.78: claude: potentially unexpected fatal signal 11.
CPU: 0 PID: 3128 Comm: claude Not tainted 6.6.87.2-microsoft-standard-WSL2 #1
# Proof it's network-related:
$ sudo unshare -n sudo -u $USER ~/.local/share/claude/versions/2.1.76 --version
2.1.76 (Claude Code)
Steps to Reproduce
- Have Claude Code installed via native installer on WSL2
- Wait for an Anthropic API outage (status.claude.com shows incident)
- Run claude --version
- Observe: Segmentation fault (core dumped)
- Run sudo unshare -n sudo -u $USER ~/.local/share/claude/versions/2.1.76 --version — works fine, proving the crash is caused by the network call
Claude Model
Opus
Is this a regression?
Yes, this worked in a previous version
Last Working Version
2.1.76 (worked the night before; same binary segfaults the next day during API outage)
Claude Code Version
2.1.78
Platform
Anthropic API
Operating System
Ubuntu/Debian Linux
Terminal/Shell
WSL (Windows Subsystem for Linux)
Additional Information
- Fresh reinstall did not fix it
- Renaming ~/.claude did not fix it
- wsl --shutdown and full Windows reboot did not fix it
- node -e "console.log('ok')" works fine
- Occurred during March 17-18 2026 API incidents on status.claude.com
- glibc 2.39, CPU has AVX/AVX2 support
This issue has 5 comments on GitHub. Read the full discussion on GitHub ↗