Sets up proper observability for Python apps when your production systems inevitably break and you need answers fast. Configures structlog for JSON logging with correlation IDs that thread through request chains, implements the four golden signals (latency, traffic, errors, saturation) with Prometheus metrics, and shows you how to instrument endpoints without creating cardinality explosions. The correlation ID propagation patterns alone will save you hours of log diving when tracking down distributed system failures. Focuses on the practical stuff that actually helps during 3am incidents.
npx skills add https://github.com/wshobson/agents --skill python-observability