What problem does it solve? Go services shipped without observability are impossible to diagnose in production. This Skill guides the instrumentation of Go applications with the five core signals — logs, metrics, traces, profiles, and RUM — so every feature is monitorable, alertable, and debuggable before it ships. ## Core Features & Use Cases - Structured Logging with slog: Set up JSON logging for production, migrate from zap/logrus/zerolog using bridge handlers, and correlate logs with traces via the otelslog bridge. - Prometheus Metrics & Alerting: Declare counters, gauges, and histograms with correct naming and bucket choices, avoid high-cardinality label traps, document metrics with PromQL-as-comments, and build multi-window burn-rate SLO alerts plus Go runtime alerts. - OpenTelemetry Tracing & Profiling: Add spans to service methods, DB queries, and external calls, propagate context correctly, control costs with sampling, and enable pprof/Pyroscope profiling via environment toggles. - Use Case: When adding a new HTTP endpoint to a Go service, use this Skill to declare latency histograms and error counters, add spans with proper error recording, emit structured context-aware logs, and wire the PromQL into Grafana dashboards and alert rules. ## Quick Start Ask the agent to add production observability — metrics, tracing, and structured logging — to your Go HTTP service.
