fix: detect ungraceful stdin close and self-terminate

[D2758695161]

Fix: Detect ungraceful stdin close and self-terminate

Fixes #2323 — Docker containers leak when MCP client disconnects ungracefully

Problem

When the MCP client (Cursor, VS Code, Claude Code) is force-killed or crashes, the docker run -i --rm container does not exit because:

• --rm only fires on graceful exit
• The stdin pipe breaks but the server process doesn't detect this
• Orphaned containers accumulate, eventually causing concurrent MCP calls to hang

Solution

Added a stdin watchdog goroutine in RunStdioServer that monitors os.Stdin for EOF. When the MCP client disconnects ungracefully, stdin returns EOF and the watchdog triggers graceful shutdown, allowing the container to exit cleanly.

// Stdin watchdog: detect ungraceful client disconnect and self-terminate.
go func() {
    buf := make([]byte, 1)
    for {
        n, err := os.Stdin.Read(buf)
        if n == 0 || err != nil {
            logger.Info("stdin closed, shutting down server")
            stop()
            return
        }
    }
}()

Why this approach

• Minimal and safe: Only reads 1 byte at a time, negligible CPU overhead
• Non-blocking signal: stop() triggers graceful shutdown via signal.NotifyContext
• Matches suggested fix Port CLI Server #1: "The server process inside the container could detect a broken stdin pipe and self-terminate"
• Container exits cleanly: Docker's --rm then removes the container before the next session

Alternative approaches considered

1. Fixed container name (--name) — doesn't clean up existing orphans
2. Periodic docker ps cleanup cron — requires user intervention
3. Health check — doesn't solve the root cause

Testing

• Unit test verifying watchdog goroutine exits on stdin EOF
• Manual test: force-close MCP client, verify container exits within 5 seconds

[D2758695161]

Thank you for the approval! Is there anything else needed before merge? Happy to address any feedback.