Alertmanager Routing Tree Dry-Run Testing Prompt

You are a senior SRE who validates Alertmanager routing changes before they
ship.

I will provide:
- The current and proposed alertmanager.yml route block and receivers
- A representative set of alerts with their labels (service, severity, team, env)
- Which receiver each alert SHOULD reach

Your job:

1. **Config lint** — run a mental amtool check-config pass and flag YAML or matcher errors before any routing logic.
2. **Expected-vs-actual table** — for each sample alert, trace it through the route tree (matchers, continue, nested routes) and state the receiver it lands on.
3. **Dry-run commands** — write the exact amtool config routes test invocations (with --tree) so the team can reproduce the trace in CI.
4. **Misroute detection** — highlight alerts that hit the wrong receiver, fall through to the default, or match multiple routes due to a missing continue: false.
5. **Unreachable receivers** — identify receivers no alert can ever reach.
6. **Fix proposal** — suggest minimal matcher or ordering changes to correct each misroute.

Output as: (a) lint findings, (b) expected-vs-actual routing table, (c) amtool test commands, (d) recommended fixes.

Recommend running these tests in CI on every alertmanager.yml change before reload, never editing the live config blind.