Skip to content

[BUG] Claude Code crashes with SIGABRT during shutdown due to MCP server termination failure #7718

@jgspratt

Description

@jgspratt

Preflight Checklist

  • I have searched existing issues and this hasn't been reported yet
  • This is a single bug report (please file separate reports for different bugs)
  • I am using the latest version of Claude Code

What's Wrong?

Claude Code crashes with SIGABRT during shutdown when attempting to terminate MCP (Model Context Protocol) servers. The crash occurs during the cleanup phase when the application cannot properly shut down connected MCP servers, leading to a Node.js assertion failure during stdio cleanup.

What Should Happen?

Claude Code should exit cleanly after properly terminating all MCP server processes and cleaning up resources.

Error Messages/Logs

SIGINT failed, sending SIGTERM to MCP server process
  Error sending SIGTERM:
  SIGTERM failed, sending SIGKILL to MCP server process
  Error sending SIGKILL:
  Error terminating process:
  Error closing client:
  Cleanup timeout reached, stopping process monitoring

  Process terminates with SIGABRT and generates core dump (261MB core file).

Steps to Reproduce

  1. Start Claude Code with MCP servers configured
  2. Use Claude Code normally (any operations)
  3. Exit Claude Code (Ctrl+C, quit command, or other exit method)
  4. Application crashes during shutdown process with core dump

Claude Model

Sonnet (default)

Is this a regression?

I don't know

Last Working Version

No response

Claude Code Version

1.0.113

Platform

Anthropic API

Operating System

Ubuntu/Debian Linux

Terminal/Shell

VS Code integrated terminal

Additional Information

Stack trace analysis:

  • Crash point: node::ResetStdio() during exit cleanup
  • Signal: SIGABRT (assertion failure)
  • Root cause: MCP servers failed to terminate (SIGINT → SIGTERM → SIGKILL all failed)
  • Node.js cleanup assertion failed when trying to reset stdio during exit

Technical details:

  • Executable: /usr/bin/node (Claude Code)
  • Core dump shows uncaught exception during shutdown sequence
  • MCP server termination timeout reached before clean exit

Metadata

Metadata

Assignees

No one assigned

    Labels

    area:corearea:mcpbugSomething isn't workinghas reproHas detailed reproduction stepsplatform:linuxIssue specifically occurs on LinuxstaleIssue is inactive

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions