What problem does it solve? Production systems fail in ways that testing never reveals: cascading failures, blocked threads, unbounded queries, and botched deployments. This Skill provides a structured framework based on Michael Nygard's Release It! to design, deploy, and operate software that survives real production conditions. ## Core Features & Use Cases - Stability Patterns & Anti-Patterns: Apply circuit breakers, bulkheads, timeouts, and retry-with-backoff while avoiding integration point failures, cascading failures, and slow-response traps. - Capacity & Deployment Guidance: Plan load/stress/soak testing, size connection and thread pools, and implement zero-downtime strategies like blue-green, canary, and feature flags. - Observability & Chaos Engineering: Design health checks, RED/USE metrics, SLOs with error budgets, and controlled failure-injection experiments with blast radius controls. - Use Case: When investigating a production outage caused by a slow downstream dependency, use this Skill to diagnose the missing timeout and absent circuit breaker, then score the system 0-10 against production-readiness criteria with concrete fixes. ## Quick Start Ask the assistant to review your service architecture for production readiness using the Release It framework and list the gaps preventing a 10/10 score.
