patrykkopycinski/o11y-service-dashboard

O11Y Service Dashboard Workflow

Guide the user through creating a comprehensive dashboard for a specific APM service.

Trigger

Use when the user asks to:

• "Create service dashboard"
• "APM dashboard"
• "Service overview"
• "Monitor service"
• "Dashboard for my service"

Also activates on keywords: "service dashboard", "APM overview", "service monitoring", "application dashboard", "service metrics"

Do NOT use when:

• "Create SLO" (→ use o11y-slo-setup)
• "Full monitoring setup" (→ use o11y-full-setup)

Steps

0. Get Cluster Context

Call get_cluster_context to get cached cluster awareness — version, health, installed features, and APM capabilities. This confirms APM data is available.

1. Identify Target Service

If the user hasn't specified a service name, call discover_o11y_data to find available APM services.

Present the list of discovered services:

• Service name, environment, language
• Throughput and data freshness

Ask the user which service they want to dashboard.

2. Discover Service Data

Call discover_o11y_data with service_names: ["<selected_service>"] to get detailed data for the target service.

3. Generate Service Summary

Call get_data_summary with the scoped discovery result.

Present the service profile:

• Transaction types and throughput
• Error rate
• Available metrics
• Suggested dashboard panels

4. User Decision Point — Dashboard Configuration

Ask the user:

• Dashboard title (suggest: "<Service Name> - Service Overview")
• Time range (suggest: last 24 hours)
• Any additional panels or customizations
• Whether to include related infrastructure or log panels (if data exists)

5. Create Dashboard

Call kibana_api with POST /api/dashboards/dashboard and the service-overview template configuration populated with the service data as the request body.

Alternative: Use create_dashboard for a higher-level approach — it generates service-overview panels automatically from the discovered data, reducing the need to manually compose panel configurations.

6. Report

Present:

• Dashboard URL (clickable link)
• List of panels created
• Suggested follow-ups: create an SLO, set up alerts, add to a composite dashboard

Tools Used

• get_cluster_context — cached cluster awareness (version, health, capabilities)
• discover_o11y_data — discover APM services and their data
• get_data_summary — generate service profile and panel suggestions
• create_dashboard — high-level dashboard creation from discovered data
• kibana_api — create the dashboard (POST /api/dashboards/dashboard)

API References

• elastic://docs/api/kibana — Kibana REST API reference for dashboard endpoints

Prerequisites

• ES_URL and ES_API_KEY configured
• APM data for at least one service

Interactive Dashboard

Interactive Dashboard: When using Claude Desktop or other ext-apps hosts, apm_service_map renders an interactive service dependency graph with throughput, latency, and error rate overlays on each node. In Cursor/CLI, it returns markdown.

Related Skills

• o11y-full-setup — complete observability setup including dashboards and SLOs
• o11y-slo-setup — create SLOs to complement service dashboards