bacoco/bmad-skills

Overview

This skill establishes instrumentation, monitoring, and alerting foundations to make systems observable and measurable. It delivers a concrete observability plan, prioritized instrumentation backlog, and SLO dashboard specifications to enable fast diagnosis and reliable alerting. The focus is on practical outputs that support launches, incident response, and ongoing reliability work.

How this skill works

I audit existing telemetry (logs, metrics, traces), tooling, and retention to document coverage and gaps. I define observability objectives tied to user journeys and business KPIs, then design a concrete instrumentation strategy (metrics taxonomy, structured logs, trace spans). Finally, I produce SLOs, dashboards, runbooks, and a prioritized implementation backlog with acceptance criteria.

When to use it

• Before a launch or major release to ensure monitoring and alerts are in place
• When logging, metrics, or tracing are missing, inconsistent, or low quality
• If you need SLOs, dashboards, or formalized runbooks for on-call teams
• When alert fatigue or noisy alerts require rationalization
• When performance or reliability work lacks trustworthy telemetry

Best practices

• Start with critical user journeys and define SLIs that map to business KPIs
• Use structured logging and consistent naming for metrics and spans
• Prioritize golden signals (latency, errors, traffic, saturation) per service
• Define guardrails for PII and retention to satisfy privacy and compliance
• Create small, verifiable instrumentation tasks with owners and test criteria

Example use cases

• Audit a microservices stack to identify telemetry blind spots and propose fixes
• Create an SLO dashboard and alert thresholds for a customer-facing API
• Define a metrics taxonomy and logging standard for a new platform
• Rationalize noisy alerts and produce a runbook and on-call expectations
• Produce a prioritized backlog to instrument payment or auth flows end-to-end