How can I add metric labels to Prometheus alert manager description?

I want to access lables of the metric which matches the expression to inform the team about which exact queue exceeds the limit:

An example of the metric

job_duration_bucket{resource_server="rs1",queue="default",connection="database",le="15"} 445
job_duration_bucket{resource_server="rs1",queue="default",connection="database",le="20"} 567
job_duration_bucket{resource_server="rs1",queue="high",connection="database",le="15"} 123
job_duration_bucket{resource_server="rs1",queue="high",connection="database",le="20"} 124

And my alerting rule

groups:
- name: alert.rules
  rules:
  - alert: JobDurationExceedance
    expr: job_duration_bucket{queue="high",le="20"} > job_duration_bucket{queue="high",le="15"}
    for: 1s
    labels:
      severity: critical
    annotations:
      { % raw % }summary: {{ $labels.resource_server }} high prio job duration exceed limit
      description: "On the {{ $labels.resource_server }} {{ $labels.queue }} prio {{ $labels.connection }} queue currently jobs exists, which are needing more than 15 minutes or more from being queued to completion."{ % endraw % }

The alert message should look like On the rs1 high prio database queue currently jobs exist...

If there is any way to achieve this I will be very thankful for advice!

I only found a way to add job labels and not metric labels

You have to quote the summary as well, then it should work fine.