fix: validate placement before sending command to prevent silent failures (BUG-001c)

The place_instance API endpoint used fire-and-forget: it sent the command
and returned HTTP 200 immediately. On a fresh cluster start, the master's
state often lacks topology/memory data, so placement raises ValueError
which was silently caught and logged. The caller never learned it failed.

Two fixes:
- API: validate placement locally before sending, return HTTP 400 on
  failure instead of silently accepting an unprocessable command
- Master: emit MetaInstancePlacementFailed on immediate placement error
  in CreateMetaInstance handler so the error surfaces in state right away

Co-Authored-By: Claude Opus 4.6 <[email protected]>

src/exo/master/api.py

@@ -312,12 +312,27 @@ class API:
         self.app.get("/v1/traces/{task_id}/raw")(self.get_trace_raw)
 
     async def place_instance(self, payload: PlaceInstanceParams):
+        model_card = await ModelCard.load(payload.model_id)
         command = PlaceInstance(
-            model_card=await ModelCard.load(payload.model_id),
+            model_card=model_card,
             sharding=payload.sharding,
             instance_meta=payload.instance_meta,
             min_nodes=payload.min_nodes,
         )
+
+        # Validate placement before sending — fail fast with a clear error
+        # instead of silently dropping the command in the master.
+        try:
+            get_instance_placements(
+                command,
+                topology=self.state.topology,
+                current_instances=self.state.instances,
+                node_memory=self.state.node_memory,
+                node_network=self.state.node_network,
+            )
+        except ValueError as exc:
+            raise HTTPException(status_code=400, detail=str(exc)) from exc
+
         await self._send(command)
 
         return CreateInstanceResponse(

src/exo/master/main.py

@@ -51,6 +51,7 @@ from exo.shared.types.events import (
     JacclSideChannelGathered,
     MetaInstanceCreated,
     MetaInstanceDeleted,
+    MetaInstancePlacementFailed,
     NodeGatheredInfo,
     TaskCreated,
     TaskDeleted,
@@ -333,6 +334,13 @@ class Master:
                                     self.state.node_network,
                                 )
                                 generated_events.extend(result.events)
+                                if result.error is not None:
+                                    generated_events.append(
+                                        MetaInstancePlacementFailed(
+                                            meta_instance_id=meta_id,
+                                            reason=result.error,
+                                        )
+                                    )
                         case DeleteMetaInstance():
                             generated_events.append(
                                 MetaInstanceDeleted(