perf: tracker injects pre-resolved route; node skips redundant tracker query

When the tracker proxies /v1/chat/completions to a first-shard node it
already holds the full registry picture. It now resolves the downstream
route inline via _select_route, strips the proxied node, and sends the
result as X-Meshnet-Route header alongside the request body.

The first-shard node reads this header in _get_remaining_route and
returns it directly, skipping the second tracker HTTP call entirely.
Falls back to the tracker query transparently when the header is absent
(direct node-to-node calls or older tracker versions).

Reduces per-inference tracker round-trips from 2 to 1.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

packages/node/meshnet_node/torch_server.py

@@ -294,18 +294,26 @@ class _TorchHandler(http.server.BaseHTTPRequestHandler):
         self._send_openai_response(result_text, model_name, stream, messages)
 
     def _get_remaining_route(self, model: str) -> list[str]:
+        # Fast path: tracker pre-resolved the downstream route and injected it as a header.
+        injected = self.headers.get("X-Meshnet-Route")
+        if injected:
+            try:
+                route = json.loads(injected)
+                if isinstance(route, list):
+                    return [str(ep) for ep in route]
+            except (json.JSONDecodeError, TypeError):
+                pass
+
+        # Slow path: query the tracker (direct node-to-node calls, or tracker didn't inject).
         server: _TorchHTTPServer = self.server  # type: ignore[assignment]
         if server.tracker_url is None:
             return []
-        # Use the backend's actual hf_repo, not the client-provided model name (which may be
-        # a lowercased or abbreviated alias that doesn't match what the tracker registered).
         route_model = getattr(server.backend, "model_id", None) or model
         try:
             url = f"{server.tracker_url}/v1/route?model={urllib.parse.quote(route_model)}"
             with urllib.request.urlopen(url, timeout=5.0) as r:
                 route_resp = json.loads(r.read())
             route = route_resp.get("route", [])
-            # Skip our own endpoint from the route (match by port so host aliases don't matter).
             own_port = server.server_address[1]
             return [ep for ep in route if not ep.rstrip("/").endswith(f":{own_port}")]
         except Exception as exc: