feat: integrate recognition engine into MQTT event pipeline

- Recognition pipeline runs between snapshot fetch and notification
- Checks recognition_enabled setting before running CLIP matching
- Known objects with quiet/suppress action skip Matrix notifications
- Vision description uses configurable backend (haiku/ollama/both)
- docker-compose.yml adds CLIP/Ollama env vars and data volume

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

docker-compose.yml

@@ -15,6 +15,12 @@ services:
       - MQTT_HOST=192.168.86.192
       - MQTT_PORT=1883
       - MATRIX_HOMESERVER=http://192.168.86.192:8008
+      - CLIP_SERVICE_URL=http://192.168.86.196:5200
+      - OLLAMA_URL=http://192.168.86.196:11434
+      - OLLAMA_MODEL=llava:13b
+      - DB_PATH=/data/frigate-notify.db
+    volumes:
+      - /mnt/docker-data/configs/frigate-notify:/data
     mem_limit: 256M
     memswap_limit: 256M
     healthcheck: