Produces an observability strategy across the three pillars — metrics, logs, traces — plus correlation, alerting philosophy, and tooling, framed around the questions the team must answer during an incident. Use when the user says "observability strategy", "instrumentation plan", "we can't debug X", "set up tracing/structured logging", or describes slow cross-service debugging. Use this to design how to SEE the system; use slo-definition to set the reliability targets the alerts defend, and incident-postmortem to analyse an incident after the fact.
