production-parity-test-designer

To design tests that catch production failures, you inventory environment gaps like DB dialects, OS behavior, and timezone/locale differences, then map each concrete failure mode to the correct unit, integration, E2E, smoke, or packaging test tier. This eliminates proxy confidence by ensuring each tier explicitly guards against specific production disparities.

CI passes while production breaks because of production parity gaps, where mocks or dev environments fail to replicate real DB dialects, dependency installation behavior, serialization, or container environments. You fix this by allocating specific test tiers to verify each environment gap.

The best way to build a PR smoke suite for database dialect parity is to inventory the specific dialect differences, map them to concrete failure modes, and propose a minimum parity smoke set with defined runtime budgets that gates pull requests against real production-equivalent verification.

To prevent historical defects from recurring, you create an adversarial regression backlog that converts past incidents into durable regression tests, explicitly mapping each historical defect to the appropriate test tier for continuous prevention.

To verify packaging integrity for production containers, you use a packaging checklist that ensures built artifacts, installs, imports, and containers truly match what production will run, verifying real production-equivalent behavior rather than just source test execution.

Production parity testing explicitly addresses CI environments that lack real dependency coverage by identifying where mocks differ from real services and assigning those environment gaps to specific integration, E2E, or smoke test tiers to eliminate false confidence.