Logging and Observability

When to use

• Wiring observability for a new service
• Debugging cross-service requests
• Removing console.log from production code

Core rules

1. Structured JSON logs in production, pretty in development
2. Every log entry has timestamp, level, message, traceId
3. No console.log in production code
4. Correlation/trace ID passed through async context
5. Health endpoint (/health) and readiness endpoint (/ready)

Reference shape (TypeScript)