feat(us-022): X-Meshnet-Start-Layer pipeline protocol for overlapping shards

When _select_route picks two nodes with overlapping registrations (e.g.
A:0-22 and B:20-24), the tracker now injects start_layer per hop so B
executes only layers 23-24, not 20-24.

- model_backend: forward_bytes + _run_layers accept start_layer offset;
  clamped to shard_start to prevent out-of-bounds indexing
- torch_server: _handle_binary_forward reads X-Meshnet-Start-Layer header;
  _run_downstream_pipeline sends it per hop; route is now list[tuple[str,int]]
- server: proxy injects {endpoint, start_layer} objects in X-Meshnet-Route;
  /v1/route response includes start_layer per node in the nodes list
- test: fake backends accept start_layer=None kwarg

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

packages/node/meshnet_node/model_backend.py

@@ -145,6 +145,7 @@ class TorchModelShard:
         shape: list[int],
         attention_mask_header: str | None,
         position_ids_header: str | None,
+        start_layer: int | None = None,
     ) -> TensorPayload | str:
         hidden_states = _tensor_from_bfloat16_bytes(body, shape, self.torch).to(
             self.device
@@ -155,7 +156,9 @@ class TorchModelShard:
         position_ids = _tensor_from_int64_header(
             position_ids_header, self.torch, self.device
         )
-        hidden_states = self._run_layers(hidden_states, attention_mask, position_ids)
+        hidden_states = self._run_layers(
+            hidden_states, attention_mask, position_ids, start_layer=start_layer
+        )
         if self.is_tail:
             return self.decode_tail(hidden_states)
         return self._payload(hidden_states, attention_mask, position_ids)
@@ -278,7 +281,21 @@ class TorchModelShard:
         )
         return dict(self.tokenizer(prompt, return_tensors="pt"))
 
-    def _run_layers(self, hidden_states: Any, attention_mask: Any, position_ids: Any) -> Any:
+    def _run_layers(
+        self,
+        hidden_states: Any,
+        attention_mask: Any,
+        position_ids: Any,
+        start_layer: int | None = None,
+    ) -> Any:
+        # start_layer overrides shard_start for overlapping-shard routing
+        # (X-Meshnet-Start-Layer header). Clamped to shard_start to prevent
+        # indexing outside the loaded weights.
+        effective_start = (
+            max(self.shard_start, start_layer)
+            if start_layer is not None
+            else self.shard_start
+        )
         position_embeddings = _rotary_position_embeddings(
             self.model,
             hidden_states,
@@ -290,7 +307,7 @@ class TorchModelShard:
             self.torch,
         )
         with self.torch.inference_mode():
-            for layer in self.layers[self.shard_start:self.shard_end + 1]:
+            for layer in self.layers[effective_start:self.shard_end + 1]:
                 hidden_states = _call_layer(
                     layer,
                     hidden_states,