Use Cases

This page is a modeling cookbook, not a feature list. Each scenario explains which model to use and why it is clearer than sending plain text notifications for everything.

Servers and NAS

Disk Alert: Message

When disk usage crosses a threshold, one strong alert is enough.

USAGE=$(df / | awk 'NR==2 { gsub("%", "", $5); print $5 }')

if [ "$USAGE" -ge 90 ]; then
  curl -X POST https://gateway.pushgo.dev/message \
    -H "Content-Type: application/json" \
    -d '{
      "channel_id": "YOUR_CHANNEL_ID",
      "password": "YOUR_CHANNEL_PASSWORD",
      "title": "Disk space warning",
      "body": "Root partition usage has reached '"$USAGE"'%.",
      "severity": "high",
      "tags": ["nas", "disk"]
    }'
fi

Backup Job: Event

A backup has a start, progress, and result. Event keeps those updates under one lifecycle.

curl -X POST https://gateway.pushgo.dev/event/create \
  -H "Content-Type: application/json" \
  -d '{
    "channel_id": "YOUR_CHANNEL_ID",
    "password": "YOUR_CHANNEL_PASSWORD",
    "title": "NAS backup",
    "status": "running",
    "message": "Backup started",
    "severity": "normal",
    "event_time": 1713750000000,
    "started_at": 1713750000000
  }'

Store the returned event_id, then call /event/update and /event/close during upload, verification, completion, or failure.

Host Status Panel: Thing

If you want to see the current state of a server over time, model the host as a Thing.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Home NAS",
  "observed_at": 1713750000000,
  "tags": ["nas", "home"],
  "attrs": {
    "online": true,
    "cpu": 0.18,
    "disk_used": 0.72,
    "temperature": 43.2
  }
}

Home Assistant and IoT

Room or Sensor: Thing

Temperature, humidity, and light are persistent state. Model each room or device as a Thing and update attrs.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Living room",
  "observed_at": 1713750000000,
  "external_ids": {
    "home_assistant": "sensor.living_room_temperature"
  },
  "location_type": "physical",
  "location_value": "home/living-room",
  "attrs": {
    "temperature": 22.5,
    "humidity": 0.46
  }
}

Door Open to Closed: Event

An open door is a process that ends when the door closes.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Balcony door open",
  "status": "open",
  "message": "The balcony door has been open for more than 5 minutes.",
  "severity": "high",
  "event_time": 1713750000000,
  "started_at": 1713750000000,
  "tags": ["home", "door"]
}

Motion Snapshot: Message

A camera motion detection with one snapshot is a Message.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Motion detected",
  "body": "Entry camera detected motion.",
  "severity": "high",
  "images": ["https://example.com/snapshot.jpg"],
  "tags": ["camera", "security"]
}

DevOps and Automation

Deployment Pipeline: Event

Deployments have lifecycle state. Create at the start, update at major stages, and close with success or failure.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Production deployment",
  "status": "building",
  "message": "Building image",
  "severity": "normal",
  "event_time": 1713750000000,
  "started_at": 1713750000000,
  "attrs": {
    "service": "api",
    "revision": "8f3c2a1"
  }
}

On failure, raise severity to critical and close the event with status=failed.

Service Health: Thing

Online state, latency, and version are current state.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "API service",
  "observed_at": 1713750000000,
  "location_type": "cloud",
  "location_value": "prod",
  "attrs": {
    "healthy": true,
    "latency_ms": 83,
    "version": "1.8.4"
  }
}

Failure Alert: Message

Even if Event records the deployment lifecycle, you can send a Message at the failure point for stronger mobile visibility.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Production deployment failed",
  "body": "The api service deployment failed. Check the pipeline logs.",
  "severity": "critical",
  "tags": ["deploy", "prod"]
}

Personal Automation

Price Monitor: Message

A price threshold is usually one alert.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Price dropped",
  "body": "The tracked product is now 199.",
  "severity": "normal",
  "url": "https://example.com/product"
}

Long Download or Transcoding Job: Event

Downloads, transcoding, and training jobs are Events because they start, update, and end.

{
  "channel_id": "YOUR_CHANNEL_ID",
  "password": "YOUR_CHANNEL_PASSWORD",
  "title": "Video transcoding",
  "status": "running",
  "message": "45% complete",
  "severity": "normal",
  "event_time": 1713750000000,
  "attrs": {
    "progress": 0.45
  }
}

Modeling Cheat Sheet

Scenario	Model	Why
Failure, success, price drop, alert	Message	One-off alert.
Deployment, backup, transcoding, incident	Event	Process with updates and an outcome.
Server, sensor, room, service	Thing	Persistent object with current state.
Alert on a specific device	Thing + Message	The entity and alert remain connected.
Incident on a service	Thing + Event	Service is the entity; incident is the lifecycle.