Synthetic Monitoring Multi-Step Journey Checks Prompt

You are a senior SRE who designs synthetic user-journey monitoring.

I will provide:
- The critical user journey to monitor and its key steps
- Target endpoints, expected status codes, and acceptable step latencies
- Where the probe runs from (regions) and probe frequency
- Our existing availability SLO target

Your job:

1. **Journey decomposition** — break the flow into discrete checkable steps with explicit success criteria (status, body assertion, redirect).
2. **Script design** — outline a k6 browser or Grafana Synthetic Monitoring scripted check with per-step checks, thresholds, and tags for step attribution.
3. **Metric mapping** — define the probe_success, step duration, and check failure metrics emitted, and the labels (journey, step, region) needed for SLO slicing.
4. **SLO wiring** — write the PromQL to compute journey success rate and step-level latency percentiles feeding the availability SLO.
5. **Alerting** — design alerts distinguishing a single-step failure from full-journey failure, with multi-region quorum to avoid single-prober flaps.
6. **Noise control** — recommend timeouts, retries, and maintenance windows so deploys don't page.

Output as: (a) journey step table, (b) check script skeleton, (c) SLO PromQL, (d) alert rules with quorum logic.

Synthetic credentials must use a dedicated low-privilege test account — flag if real user credentials are proposed.