Agent Skills: Alert Management

Alert Management

Table of Contents

• Overview
• When to Use
• Quick Start
• Reference Guides
• Best Practices

Overview

Design and implement sophisticated alert management systems with PagerDuty integration, escalation policies, alert routing, and incident coordination.

When to Use

• Setting up alert routing
• Managing on-call schedules
• Coordinating incident response
• Creating escalation policies
• Integrating alerting systems

Quick Start

Minimal working example:

// pagerduty-client.js
const axios = require("axios");

class PagerDutyClient {
  constructor(apiToken) {
    this.apiToken = apiToken;
    this.baseUrl = "https://api.pagerduty.com";
    this.eventUrl = "https://events.pagerduty.com/v2/enqueue";

    this.client = axios.create({
      baseURL: this.baseUrl,
      headers: {
        Authorization: `Token token=${apiToken}`,
        Accept: "application/vnd.pagerduty+json;version=2",
      },
    });
  }

  async triggerEvent(config) {
    const event = {
      routing_key: config.routingKey,
      event_action: config.eventAction || "trigger",
      dedup_key: config.dedupKey || `event-${Date.now()}`,
      payload: {
        summary: config.summary,
// ... (see reference guides for full implementation)

Reference Guides

Detailed implementations in the references/ directory:

| Guide | Contents | |---|---| | PagerDuty Client Integration | PagerDuty Client Integration | | Alertmanager Configuration | Alertmanager Configuration | | Alert Handler Middleware | Alert Handler Middleware | | Alert Routing Engine | Alert Routing Engine | | Docker Compose Alert Stack | Docker Compose Alert Stack |

Best Practices

✅ DO

• Set appropriate thresholds
• Implement alert deduplication
• Use clear alert names
• Include runbook links
• Configure escalation properly
• Test alert rules
• Monitor alert quality
• Set repeat intervals
• Track alert metrics
• Document alert meanings

❌ DON'T

• Alert on every anomaly
• Ignore alert fatigue
• Set thresholds arbitrarily
• Skip runbooks
• Alert without action
• Disable alerts in production
• Use vague alert names
• Forget escalation policies
• Re-alert too frequently