Prometheus
App · 8 actions
Query your Prometheus metrics with PromQL, now or over a time range, and check its alerts, rules and scrape targets. Start a workflow when an alert starts firing.
Actions
| Action | What it does |
|---|---|
| Find series | List the series that match a selector, with all their labels, optionally only those seen within a time range. |
| Get label values | List the values a label takes, such as every job or instance, optionally only on series matching a selector. |
| Get metric info | Get a metric's type (counter, gauge, histogram or summary), unit and help text, or list them for every metric. |
| List alerts | List the alerts firing or pending now, with their labels, annotations and when they became active. |
| List rules | List the server's alerting or recording rules with their query, health and, for alerts, whether they are firing. |
| List targets | List what Prometheus scrapes and whether each target is up, with the last error for any that are down. |
| Run query | Run a PromQL query for one moment, now unless you pick another, and get each result's labels and value. |
| Run range query | Run a PromQL query over a time range, one point per step, and get each series' points. |
Find series
List the series that match a selector, with all their labels, optionally only those seen within a time range.
Inputs
- Selector
matchtextRequired - A series selector, such as up{job="node"}.
- Seen after
startdate and time - Only series with data after this moment.
- Seen before
enddate and time - Only series with data before this moment.
- Limit
limitnumber - The most to return. Defaults to 100.From 1 to 1000.
Outputs
countnumber- How many series were returned
has_moretrue/false- True when more series match than the limit
matchtext- The selector
serieslist- The matching series' labels
Example output
{
"count": 1,
"has_more": false,
"match": "up",
"series": [
{
"__name__": "up",
"instance": "localhost:9090",
"job": "prometheus"
}
]
}Get label values
List the values a label takes, such as every job or instance, optionally only on series matching a selector.
Inputs
- Label
labeltextRequired - The label, such as job or instance.
- Only on series
matchtext - A series selector, such as up{job="node"}. Leave empty for all series.
- Limit
limitnumber - The most to return. Defaults to 100.From 1 to 1000.
Outputs
countnumber- How many values were returned
has_moretrue/false- True when it has more values than the limit
labeltext- The label
valueslist- Its values, in order
Example output
{
"count": 2,
"has_more": false,
"label": "job",
"values": [
"node",
"prometheus"
]
}Get metric info
Get a metric's type (counter, gauge, histogram or summary), unit and help text, or list them for every metric.
Inputs
- Metric
metrictext - The metric. Leave empty to list every metric.
- Limit
limitnumber - The most to return. Defaults to 100.From 1 to 1000.
Outputs
countnumber- How many metrics were returned
has_moretrue/false- True when there are more than the limit
metricslist- The metrics
Example output
{
"count": 1,
"has_more": false,
"metrics": [
{
"help": "Whether the target is up",
"name": "up",
"type": "gauge",
"unit": ""
}
]
}List alerts
List the alerts firing or pending now, with their labels, annotations and when they became active. Firing alerts come first.
Inputs
- State
statetext - Which alerts to include. Defaults to firing or pending.One of:
any(Firing or pending),firing,pending. Default:any. - Alert
alert_nametext - Only alerts of this name. Leave empty for all.
Outputs
alertslist- The alerts
countnumber- How many alerts were returned
firing_countnumber- How many of them are firing
Example output
{
"alerts": [
{
"active_since": "2026-10-27T10:15:00Z",
"annotations": {
"summary": "More than 5% of requests are failing"
},
"labels": {
"alertname": "HighErrorRate",
"job": "api",
"severity": "critical"
},
"name": "HighErrorRate",
"severity": "critical",
"state": "firing",
"value": 0.072
}
],
"count": 1,
"firing_count": 1
}List rules
List the server's alerting or recording rules with their query, health and, for alerts, whether they are firing.
Inputs
- Rules
rule_typetext - Which rules to list. Defaults to alerting rules.One of:
alert(Alerting rules),record(Recording rules),any(Both). Default:alert. - Group
grouptext - Only rules in this group. Leave empty for every group.
Outputs
countnumber- How many rules
ruleslist- The rules
Example output
{
"count": 1,
"rules": [
{
"active_alert_count": 1,
"group": "api",
"health": "ok",
"labels": {
"severity": "critical"
},
"last_error": "",
"name": "HighErrorRate",
"query": "rate(errors_total[5m]) > 0.05",
"state": "firing",
"type": "alert"
}
]
}List targets
List what Prometheus scrapes and whether each target is up, with the last error for any that are down. Down targets come first.
Inputs
- Health
healthtext - Only targets in this state. Defaults to any.One of:
any,up,down,unknown. Default:any. - Job
jobtext - Only targets of this job. Leave empty for every job.
Outputs
countnumber- How many targets were returned
down_countnumber- How many of them are down
targetslist- The targets
Example output
{
"count": 1,
"down_count": 1,
"targets": [
{
"health": "down",
"instance": "db-1:9100",
"job": "node",
"labels": {
"instance": "db-1:9100",
"job": "node"
},
"last_error": "connection refused",
"last_scrape": "2026-10-27T10:19:45Z",
"scrape_url": "http://db-1:9100/metrics"
}
]
}Run query
Run a PromQL query for one moment, now unless you pick another, and get each result's labels and value. 'value' holds the first result's number, handy for a single total or rate.
Inputs
- Query
querytextRequired - A PromQL expression, such as sum(rate(http_requests_total[5m])).
- At
at_timedate and time - The moment to evaluate it at. Leave empty for now.
Outputs
countnumber- How many results there are
querytext- The query that ran
result_typetext- vector, scalar, string or matrix
resultslist- Each result with its labels and value
truncatedtrue/false- True when there were more than 1000 results
valuenumber- The first result's value; empty when there are none
Example output
{
"count": 1,
"query": "up",
"result_type": "vector",
"results": [
{
"labels": {
"__name__": "up",
"instance": "localhost:9090",
"job": "prometheus"
},
"raw_value": "1",
"time": "2026-10-27T10:20:00Z",
"value": 1
}
],
"truncated": false,
"value": 1
}Run range query
Run a PromQL query over a time range, one point per step, and get each series' points. For trends and reports.
Inputs
- Query
querytextRequired - A PromQL expression, such as sum(rate(http_requests_total[5m])).
- From
startdate and timeRequired - When the range begins.
- To
enddate and time - When the range ends. Leave empty for now.
- Resolution
steptext - How far apart the points are. Defaults to every minute. A duration such as 30s or 2h works too.One of:
15s(Every 15 seconds),1m(Every minute),5m(Every 5 minutes),15m(Every 15 minutes),1h(Every hour),6h(Every 6 hours),1d(Every day). Default:1m.
Outputs
countnumber- How many series there are
endtext- When it ended, in UTC (ISO 8601)
querytext- The query that ran
serieslist- Each series with its points
starttext- When the range began, in UTC (ISO 8601)
step_secondsnumber- Seconds between points
truncatedtrue/false- True when there were more than 500 series
Example output
{
"count": 1,
"end": "2026-10-27T10:02:00Z",
"query": "up",
"series": [
{
"labels": {
"__name__": "up",
"instance": "localhost:9090",
"job": "prometheus"
},
"points": [
{
"raw_value": "1",
"time": "2026-10-27T10:00:00Z",
"value": 1
},
{
"raw_value": "1",
"time": "2026-10-27T10:01:00Z",
"value": 1
},
{
"raw_value": "1",
"time": "2026-10-27T10:02:00Z",
"value": 1
}
]
}
],
"start": "2026-10-27T10:00:00Z",
"step_seconds": 60,
"truncated": false
}