dashboard-builder

affaan-m/everything-claude-code

Build operational dashboards that answer real questions, not vanity boards.

What is dashboard-builder?

Creates monitoring dashboards for Grafana, SigNoz, and similar platforms focused on answering operator questions: is it healthy, where's the bottleneck, what changed, what action to take. Use when turning metrics into a working dashboard instead of displaying every available metric.

Start from operator questions rather than visual layout
Organize panels around health, latency, throughput, saturation, and service-specific risks
Study target platform schema and existing dashboards before building
Build minimum useful board with overview, performance, resources, and service-specific sections
Remove vanity panels that don't answer real operational questions
Ensure valid JSON, clear grouping, titles, units, meaningful thresholds, and sensible defaults

How to install dashboard-builder

npx skills add https://github.com/affaan-m/everything-claude-code --skill dashboard-builder

Claude Code

Cursor

Windsurf

Cline

How to use dashboard-builder

1.Define the operating questions your dashboard must answer (health, latency, throughput, saturation, service risks)
2.Study the target platform's schema by inspecting existing dashboards for JSON structure and query language
3.Build the minimum useful board with four sections: overview, performance, resources, and service-specific panels
4.Add titles, units, and meaningful thresholds to each panel
5.Configure variables for common filters and set sensible default time range and refresh rates
6.Review and remove any panels that don't answer a real operational question

Use cases

Good for

Build a Kafka monitoring dashboard with broker count, under-replicated partitions, consumer lag, and disk pressure
Create a Grafana dashboard for Elasticsearch with cluster health, shard allocation, search latency, and JVM metrics
Make a SigNoz dashboard for an API gateway with request rate, latency percentiles, error rate, and upstream health
Turn a metrics list into an operational dashboard with proper structure and filtering variables
Monitor service health and performance with panels grouped by availability, throughput, and resource utilization

Who it's for

Site reliability engineers (SREs) building operational dashboards
DevOps engineers creating monitoring solutions for production systems
Backend engineers setting up service-specific monitoring
Platform operators needing to diagnose issues quickly

dashboard-builder FAQ

Should I include every available metric in the dashboard?

No. Every panel should answer a real operator question. If a metric doesn't help answer 'is it healthy?', 'where's the bottleneck?', 'what changed?', or 'what action to take?', remove it.

What structure should I use for organizing panels?

Organize around operator questions: health/availability, latency/performance, throughput/volume, saturation/resources, and service-specific risks. A recommended layout is overview, performance, resources, then service-specific sections.

How do I start building a dashboard for a specific service?

Start by studying existing dashboards for your platform to understand the JSON structure and query language. Then define the operating questions specific to that service before designing any visual layout.

What makes a panel valuable versus vanity?

A valuable panel has a clear title, proper units, meaningful thresholds or status colors, and answers a specific operational question. Vanity panels show metrics just because they exist without actionable insight.

Which platforms does this skill support?

It works with Grafana, SigNoz, and similar monitoring platforms that use JSON-based dashboard definitions and support variables and threshold styling.

Full instructions (SKILL.md)

Source of truth, from affaan-m/everything-claude-code.