deployment-procedures

A safe production release workflow requires structured stages including pre-deployment verification checks, staged rollout execution, and post-deployment monitoring to ensure system stability. This skill provides comprehensive guidelines and decision checklists to walk you through each phase safely.

Zero-downtime deployment is best achieved using rolling, blue-green, or canary release strategies to gradually shift traffic without interrupting service. The skill discusses these strategies in detail to help you select and implement the right approach for your platform.

To prepare and execute a rollback, you must establish rollback principles and verify system backups before deployment begins. This skill outlines rollback strategies and emergency procedures to revert changes quickly when post-deployment verification detects failures.

When a service is down after a production release, you should follow emergency procedures that prioritize immediate investigation and rollback actions. The skill guides you through the priority actions and investigation orders required to restore functionality quickly.

Common deployment anti-patterns include skipping pre-deployment verification, large infrequent releases, and ignoring post-deployment monitoring. This skill highlights these pitfalls and provides best practices, such as using feature flags, to ensure safe production workflows.

You do not need specific platform tools, as the skill provides general deployment principles, decision trees for platform selection, and verification checklists applicable across environments. It uses a Vercel deployment example to illustrate the workflow.