**TL;DR** - Design your observability stack before you need it. Not after.
1. Describe your architecture: services, SLO targets, data flows
2. Review the generated observability plan
3. Apply the instrumentation recommendations
**Example Prompt:**
Design an observability strategy for our e-commerce platform. SLO: 99.9% uptime, p99 checkout latency under 2s.
| Pros | Cons |
|---|---|
| Ensures observability is designed in, not bolted on | Generated alerts need tuning |
| Catches gaps before they become blind spots | Requires someone who knows the system |
|---|
| Aligns observability with business SLOs | Can't fully capture "why" behind SLO choices |
|---|