Runbooks and observability

The operational muscle teams lean on at 2am. Key points:

• Runbooks are versioned, reviewed procedures. Steps that touch production pause for a human approval through the same gate the rest of the platform uses, and every run is logged, so the next on-call engineer inherits a record.
• Observability reads from the metrics, logs, and traces you already collect. IntegraCI renders rate, errors, and duration per service. There is no second pipeline to run; you point at the data you already have.
• Both sit next to the service in the catalog, so the answer is beside the work instead of in a separate tab.