Prometheus Agent

Required Parameters

• Language Model: The AI model for processing
• Endpoint: Prometheus server URL
• Authentication: Authentication credentials
• Query Type: Type of metric query
• Time Range: Default time range

Optional Configuration

• Query Options: Query settings
  • timeout: Query timeout
  • maxPoints: Maximum data points
  • step: Query resolution step
• Alerts: Alert configuration
  • rules: Alert rules
  • channels: Notification channels
  • templates: Alert templates
• Analysis: Analysis options
  • aggregation: Aggregation methods
  • prediction: Prediction settings
  • correlation: Correlation analysis

Output Format

{
  "result": {
    "query": {
      "promql": string,
      "parameters": object
    },
    "data": {
      "resultType": string,
      "result": array,
      "stats": object
    },
    "analysis": {
      "summary": string,
      "insights": array,
      "anomalies": array
    }
  }
}

Example Usage

const prometheusAgent = new PrometheusAgent({
  languageModel: "gpt-4",
  endpoint: "http://prometheus.example.com:9090",
  authentication: {
    type: "bearer",
    token: process.env.PROMETHEUS_TOKEN
  },
  queryType: "range",
  timeRange: {
    start: "-1h",
    end: "now"
  },
  queryOptions: {
    timeout: 30,
    maxPoints: 1000,
    step: "15s"
  },
  alerts: {
    rules: [
      {
        name: "HighLatency",
        condition: "http_request_duration_seconds > 1",
        duration: "5m"
      }
    ],
    channels: ["slack", "email"],
    templates: {
      default: "{{ .Alert.Name }}: {{ .Alert.Description }}"
    }
  },
  analysis: {
    aggregation: ["sum", "avg", "max"],
    prediction: {
      enabled: true,
      window: "1h"
    },
    correlation: {
      enabled: true,
      metrics: ["cpu", "memory", "latency"]
    }
  }
});

const result = await prometheusAgent.process({
  input: "Show CPU usage trends for all web servers over the last hour"
});