[BUG] Claude crushed my linux box by allocating more than 100% swap

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?

Then when I managed to kill it from the CLI, /resume immediately repeated the behaviour.

What Should Happen?

It should use an appropriate amount of ram and self regulate as soon as it realises its about to bring down the host.

Error Messages/Logs

top - 19:28:34 up 24 days,  8:31,  9 users,  load average: 3.19, 20.32, 16.40
Tasks: 446 total,   3 running, 434 sleeping,   0 stopped,   9 zombie
%Cpu(s):  7.3 us,  1.1 sy,  0.0 ni, 91.6 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
MiB Mem :  63427.6 total,  31274.0 free,  31182.4 used,   1787.6 buff/cache
MiB Swap:  98304.0 total,  95014.8 free,   3289.2 used.  32245.2 avail Mem

    PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
3480863 john      20   0  144.9g  22.9g  52800 R 100.3  36.9   2:08.90 claude
3483650 john      20   0   22.1g   2.5g  53760 R 100.0   4.1   0:22.77 claude
   1795 neo4j     20   0   23.7g  39824  12564 S   0.3   0.1  98:03.70 java
3473234 john      20   0   62.3g 545068  30528 S   0.3   0.8   0:36.51 node
      1 root      20   0   22780   8448   6720 S   0.0   0.0   0:18.72 systemd

Steps to Reproduce

Unknown.

But its completely trashed now:
john@nyx:~/elspeth$ claude --version
Aborted (core dumped)
john@nyx:~/elspeth$ claude --version
Aborted (core dumped)
john@nyx:~/elspeth$

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

:atest/no way to know.

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Xterm

Additional Information

_No response_