What problem does it solve? Teams often ship based on gut feeling rather than evidence, leading to production incidents, untested rollback plans, and chaotic Friday-night firefighting. This Skill turns "ready to ship" into measurable criteria: auditable go/no-go checklists, fast smoke test suites, metric-gated staged rollouts, and pre-defined rollback thresholds. ## Core Features & Use Cases - Go/No-Go Checklists: Automated and manual verification items (CI green, P0/P1 bugs closed, performance budgets, security scans) stored as versioned, sign-off artifacts. - Smoke Test Suite Design: Guidance for building a sub-5-minute suite covering critical user journeys, with environment-specific variants for staging, production, and post-deployment. - Staged Rollout & Rollback Automation: Canary-to-100% promotion criteria relative to baseline metrics, automated rollback triggers, kill-switch patterns, and feature-flag rollout tooling comparisons (LaunchDarkly, Statsig, GrowthBook, Unleash, Flagsmith). - Use Case: Before deploying v2.4, run the checklist, confirm smoke tests pass on staging, define rollback triggers (e.g., error rate >2x baseline for 5 minutes), roll out to 1% of traffic, and promote only when error, latency, and business metrics stay within baseline. ## Quick Start Ask the AI to build a go/no-go release checklist and rollback plan for your upcoming deployment, including smoke tests and staged rollout criteria.
