Observability & Errors

Overview

You can't fix what you can't see. A production SaaS needs three signals: error tracking (Sentry) for exceptions with stack + context, structured logs (JSON with a correlation id) for tracing a request, and product analytics (PostHog) for what users actually do. Add a global error boundary, attach tenant/user/request context to every signal, and scrub secrets and PII before anything leaves the process.

When to use

• Standing up a new app's logging/monitoring.
• Debugging issues you can't reproduce locally.
• You need to know which features are used or where users drop off.

Structured logging with context

// src/server/log.ts
import "server-only";