fix(core): gracefully stop FXServer child on shutdown signals

[FjamZoo]

Description

Previously the core only spawned and managed the FXServer child but never handled its own termination. On SIGTERM (docker stop, systemd, etc) or SIGINT (Ctrl+C), the panel process was killed immediately and the spawned FXServer was left to be torn down abruptly rather than stopped cleanly.

This adds SIGTERM/SIGINT handlers in apps/core/src/index.ts that:

• call ProcessManager.stop() to stop the running FXServer child,
• close the Fastify server,
• then process.exit(0),
• with a shuttingDown guard so a repeated/duplicate signal is a no-op.

ProcessManager.onExit already treats exit code 143 as a clean stop, so no spurious "crashed" state is reported.

---

Tests

• Tested in development mode
• Passes typecheck
• Builds & runs on Windows
• Builds & runs on Linux

---

Additional Notes

Verified by building the Linux target and running the production binary in a container, with the server in the running state:

• SIGTERM (docker stop): child stopped gracefully (pm.stop() ran), panel exited with code 0 in rougly 2-3 seconds.
• SIGINT (docker kill -s SIGINT): same clean shutdown, exit code 0.
• Restart: server starts again correctly after the process comes back up.

[Maximus7474]

Seems to hang on windows (prod build) on the fastify.close() call.
It's been a couple minutes and only process manager has been stopped.

Testing code:

	try {
		await pm.stop();
		console.log('stopped process manager')
		await fastify.close();
		console.log('closed fastify server')
	} catch (err) {

[Maximus7474]

As @andreutu says: LGTM