llama-cpp: pause xmrig during active inference requests

Add sidecar service that polls llama-cpp /slots endpoint every 3s.
When any slot is processing, stops xmrig. Restarts xmrig after 10s
grace period when all slots are idle. Handles unreachable llama-cpp
gracefully (leaves xmrig untouched).

configuration.nix

@@ -65,6 +65,8 @@
     ./services/p2pool.nix
     ./services/xmrig.nix
 
+    ./services/llama-cpp-xmrig-pause.nix
+
     # KEEP UNTIL 2028
     ./services/caddy_senior_project.nix