python-observability

python-observability is an agent skill from the wshobson/agents collection for instrumenting Python applications with structured logs, metrics and distributed traces, so that production problems can be understood without deploying new code. The goal is to answer what happened, where, and why when something breaks. The skill covers structured logging, which emits logs as JSON with consistent fields so they can be searched and alerted on, while noting that a human-readable format may suit local development better. It covers metrics collection with Prometheus, distributed tracing across services, and propagating correlation IDs through a chain of requests, so one request can be followed across every service it touches. It applies when adding structured logging to an application, implementing metrics collection, setting up distributed tracing, propagating correlation IDs, debugging production issues, and building observability dashboards. The skill is written for engineers who run Python services in production and need their telemetry to be useful during an incident. The agent adds the instrumentation in code, and the team chooses the backends that receive the logs, metrics and traces. The skill is installed with the skills CLI from the wshobson/agents repository. Its skills.sh page showed about 11.3K installs and passing results from three security audits at the time of the scan.…