Scenario Guide¶

Guide Maps¶

graph LR
  setup["Policy setup"] --> rules["Register rules"]
  rules --> activate["Activate rules"]
  activate --> observe["Observe samples"]
  observe --> alerts["Publish alerts"]

flowchart TD
  inputs["Read the fixed inputs"] --> actions["Follow the scenario actions"]
  actions --> outputs["Compare the expected outputs"]
  outputs --> review["Use bundles, code, and tests to explain why"]

Use this guide when you want the exact fixed example behind make demo, make tour, and the CLI inspection routes. The goal is to keep the capstone's teaching scenario stable and reviewable instead of hiding it in implementation details.

Fixed policy¶

policy id: service-monitoring
monitored metric: cpu

Fixed rule registrations¶

Rule id	Mode	Threshold	Window	Severity	Meaning
`cpu-hot`	`threshold`	`0.9`	`1`	`critical`	fire when the latest sample breaches the hot threshold
`cpu-sustained`	`consecutive`	`0.8`	`3`	`warning`	fire when the latest three samples all stay above the sustained threshold

Fixed samples¶

Timestamp	Metric	Value	Why it matters
`2026-04-02T09:00:00`	`cpu`	`0.82`	starts the sustained-evaluation window without triggering the hot rule
`2026-04-02T09:01:00`	`cpu`	`0.93`	breaches the hot threshold
`2026-04-02T09:02:00`	`cpu`	`0.95`	completes the sustained window and triggers both active rules

Expected outcomes¶

Surface	Expected outcome
cycle report	`alerts_published = 2`
active rules	`cpu-hot`, `cpu-sustained`
open incidents	one for each active rule
incident history	two entries under the `cpu` metric

Best follow-up surfaces¶

read TOUR.md when you want the same scenario as a narrative route
read DOMAIN_GUIDE.md when the terms are still fuzzier than the sequence
run make inspect when you want the saved outputs for this exact scenario