Production Readiness

Best practices for production-grade code

Default

Language / Topic

Nix

Rules

Production Readiness · balanced balanced

- Handle edge cases and failures gracefully — never assume the happy path.

- Add structured logging for all critical operations and error states.

- Use health check endpoints (`/healthz`, `/readyz`) for load balancers and orchestrators to verify service availability.

- Validate all external inputs at system boundaries. Trust nothing from users, APIs, or files.

- Add health check endpoints and readiness probes for all services.

- Use feature flags for risky rollouts — deploy dark, enable incrementally.

- Write code that fails loudly in development and recovers gracefully in production.