fix: reduce MoE concurrency 2→1 to prevent thermal timeout (94°C)

Strix Halo running qwen3.6-35B-A3B was hitting 94°C with 2 concurrent
slots, causing 300s request timeouts. Mumuni + Koby accumulated 15
timeouts in the last hour. Reduced to 1 slot for thermal headroom.

Medium and Default tiers already route VLM before MoE as fallback,
minimizing overflow traffic to the hot GPU.

router/router.py

@@ -19,7 +19,7 @@ GPU_URLS = {
 }
 # Max concurrent requests per GPU (based on llama.cpp --parallel)
 GPU_MAX_CONCURRENT = {
-    "qwen3.6-35B-A3B": 2,   # 2 slots
+    "qwen3.6-35B-A3B": 1,   # 1 slot (thermal management: 94C at 2 concurrent)
     "qwen3.6-27B-code": 1,  # 1 slot (24GB VRAM saturated at 256K ctx)
     "qwen3.5-9b-vlm": 2,       # 2 slots (12GB VRAM, 4GB headroom)
 }
@@ -229,7 +229,7 @@ def route(rd, tier):
     
     # TIER 3: Medium complexity — Dense primary (speed), MoE fallback
     if t <= 25000 and turns <= 15:
-        candidates = [m for m in ["qwen3.6-27B-code","qwen3.6-35B-A3B","qwen3.5-9b-vlm"] if m in avail]
+        candidates = [m for m in ["qwen3.6-27B-code","qwen3.5-9b-vlm","qwen3.6-35B-A3B"] if m in avail]
         result = select_best_gpu(candidates, "medium")
         if result: return result
     
@@ -240,7 +240,7 @@ def route(rd, tier):
         if result: return result
     
     # TIER 5: Default — balanced distribution: Dense first (speed), MoE second (capacity)
-    candidates = [m for m in ["qwen3.6-27B-code","qwen3.6-35B-A3B","qwen3.5-9b-vlm"] if m in avail]
+    candidates = [m for m in ["qwen3.6-27B-code","qwen3.5-9b-vlm","qwen3.6-35B-A3B"] if m in avail]
     result = select_best_gpu(candidates, "default")
     if result: return result
     return {"model":avail[0],"reason":"last_resort"}