feat: MLX /status API 통합 — GPU 무접촉 부하 측정

- model_adapter.status_check(): GET /status, 미지원 백엔드는 None
- backend_registry.get_load_status(): /status API 1순위, hybrid fallback
  - active_jobs >=4 매우 바쁨, >=2 바쁨, =1 보통, =0 여유
  - source: status_api | hybrid 표시
- worker._build_system_status(): source별 분기, active_jobs/total_requests 표시
- hybrid 경로의 self-job 제외 (active > 1)
- health loop에서 status도 같이 캐시 (지원 백엔드만)

Mac mini /opt/mlx-proxy.py에 active_jobs 카운터 + /status 엔드포인트 추가됨

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

nanoclaude/services/worker.py

@@ -100,19 +100,26 @@ async def _build_system_status(force_measure: bool = True) -> str:
         load = await backend_registry.get_load_status(role, force_measure=force_measure)
         connected = "정상" if info["healthy"] else "연결 안 됨"
         load_str = load["load"]
-
-        baseline = load["health_baseline_ms"]
-        ratio = (load["health_ms"] / baseline) if baseline > 0 else 1.0
+        source = load.get("source", "hybrid")
 
         line1 = f"{info['name']}: {connected} — {load_str}"
-        line2 = f"  health {load['health_ms']:.0f}ms (baseline {baseline:.0f}ms, {ratio:.1f}배)"
-        if load["measured"]:
-            line3 = f"  ping {load['inference_ms']:.0f}ms"
-        else:
-            line3 = "  ping: 측정 안 함"
         lines.append(line1)
-        lines.append(line2)
-        lines.append(line3)
+
+        if source == "status_api":
+            # /status API 직접 조회 — 정확한 active_jobs
+            active = load.get("active_jobs", 0)
+            total = load.get("total_requests", 0)
+            lines.append(f"  active_jobs: {active} (total: {total}) [status API]")
+            lines.append(f"  health {load['health_ms']:.0f}ms")
+        else:
+            # hybrid fallback (health baseline + 조건부 ping)
+            baseline = load["health_baseline_ms"]
+            ratio = (load["health_ms"] / baseline) if baseline > 0 else 1.0
+            lines.append(f"  health {load['health_ms']:.0f}ms (baseline {baseline:.0f}ms, {ratio:.1f}배)")
+            if load["measured"]:
+                lines.append(f"  ping {load['inference_ms']:.0f}ms")
+            else:
+                lines.append("  ping: 측정 안 함")
         lines.append("")
 
     queue = jq_module.job_queue.stats if jq_module.job_queue else {"pending": 0, "active": 0}