Alerts are not closed at Opsgenie receiver

What did you do?
I am using AlertManager to send alerts to Opsgenie. When the alert is resolved on backend (Prometheus/Alertmanager), I just receive another email about the alert but it remains open in Opsgenie.

What did you expect to see?
Alert being marked as closed in Opsgenie when it gets resolved in Prometheus

What did you see instead? Under which circumstances?
I just get notified by email about the alert, just like in the case when alert gets created.

Environment

• System information:
  Darwin 22.4.0 arm64

• Alertmanager version:
  v0.25.0

• Prometheus version:
  v2.37.2

• Alertmanager configuration file:

route:
  receiver: opsgenie-dispatcher
  routes:
  - matchers:
      - send_to=opsgenie
    receiver: opsgenie-dispatcher
    repeat_interval: 6h
    group_interval: 1m
    group_wait: 1m
    group_by: ['team', 'environment', 'alertname']

receivers:
- name: opsgenie-dispatcher
  opsgenie_configs:
  - send_resolved: true
    api_url: https://api.opsgenie.com
    api_key: <API_KEY>
    message: '{{ .CommonLabels.alertname }}'
    description: '{{ template "opsgenie.description" . }}'
    priority: '{{ .CommonLabels.priority }}'
    responders:
    - type: team
      name: '{{ .CommonLabels.team }}'

templates:
  - ./template/*.tmpl

• Prometheus configuration file:

global:
  scrape_interval: 10s
  scrape_timeout: 10s

rule_files:
  - alert.yml

alerting:
  alertmanagers:
    - scheme: http
      static_configs:
        - targets: [ 'alertmanager:9093' ]

scrape_configs:
  - job_name: services
    metrics_path: /metrics
    static_configs:
      - targets:
          - 'prometheus:9090'
          - 'idonotexists:564'

alert.yml

groups:
  - name: DemoAlerts
    rules:
      - alert: InstanceDown 
        expr: up{job="services"} == 0 
        for: 0m
        labels:
          send_to: opsgenie
          priority: "P2"
          team: test-prometheus-integration
        annotations:
          description: "Instance {{$labels.instance}} is down for over 1m"

• Logs:
  When alert is created:

ts=2023-06-06T10:03:16.884Z caller=dispatch.go:163 level=debug component=dispatcher msg="Received alert" alert=InstanceDown[c2e4e72][active]
ts=2023-06-06T10:04:16.884Z caller=dispatch.go:515 level=debug component=dispatcher aggrGroup="{}/{send_to=\"opsgenie\"}:{alertname=\"InstanceDown\", team=\"test-prometheus-integration\"}" msg=flushing alerts=[InstanceDown[c2e4e72][active]]
ts=2023-06-06T10:04:16.885Z caller=opsgenie.go:139 level=debug integration=opsgenie alert="{}/{send_to=\"opsgenie\"}:{alertname=\"InstanceDown\", team=\"test-prometheus-integration\"}"
ts=2023-06-06T10:04:17.315Z caller=notify.go:743 level=debug component=dispatcher receiver=opsgenie-dispatcher integration=opsgenie[0] msg="Notify success" attempts=1

When alert is resolved:

ts=2023-06-06T10:08:16.881Z caller=dispatch.go:163 level=debug component=dispatcher msg="Received alert" alert=InstanceDown[c2e4e72][resolved]
ts=2023-06-06T10:08:16.885Z caller=dispatch.go:515 level=debug component=dispatcher aggrGroup="{}/{send_to=\"opsgenie\"}:{alertname=\"InstanceDown\", team=\"test-prometheus-integration\"}" msg=flushing alerts=[InstanceDown[c2e4e72][resolved]]
ts=2023-06-06T10:08:16.885Z caller=opsgenie.go:139 level=debug integration=opsgenie alert="{}/{send_to=\"opsgenie\"}:{alertname=\"InstanceDown\", team=\"test-prometheus-integration\"}"
ts=2023-06-06T10:08:17.069Z caller=notify.go:743 level=debug component=dispatcher receiver=opsgenie-dispatcher integration=opsgenie[0] msg="Notify success" attempts=1

Start with the Opsgenie integration entry point named in the logs, opsgenie.go, and reproduce the report using the supplied Alertmanager configuration and resolved-alert logs. Trace how resolved notifications are sent and verify that the corresponding Opsgenie alert is marked closed rather than receiving only another email.