03: Scenario Pack - Standardized Technical Components
Why This Pack Exists
Skill 1.1.3 is about turning one successful design into a repeatable platform. These scenarios focus on what happens when teams solve the same problem five different ways and drift slowly turns into operational risk.
This pack is especially useful for platform engineers, Tech Leads, and architects who want to standardize:
- Bedrock client behavior,
- prompt storage and promotion,
- telemetry naming,
- message contracts,
- IAM patterns.
How To Use The Pack
- Start with the scenario closest to a platform concern in your environment.
- Identify which component should have been shared and versioned.
- Review the runbook and questions in the deep-dive file.
- Translate the fix into a reusable package, template, or control rather than a one-off patch.
Scenario Index
| Scenario | What breaks | Standardization focus | Scenario file |
|---|---|---|---|
| 1. Bedrock client config drift | Services behave differently during the same upstream event | Shared client factory and config ownership | scenarios/01-bedrock-client-config-drift.md |
| 2. Prompt template drift | Dev, staging, and prod behave differently without code changes | Prompt registry, versioning, and deployment discipline | scenarios/02-prompt-template-drift.md |
| 3. Metric naming inconsistency | Dashboards and alarms tell the wrong story | Shared telemetry schema and metric library | scenarios/03-cloudwatch-metric-naming-drift.md |
| 4. Missing standard error envelope | Downstream consumers crash on shape changes | Contract-first response envelopes | scenarios/04-standard-error-envelope.md |
| 5. IAM role pattern drift | One environment works while another fails or becomes over-permissive | Reusable least-privilege IAM modules | scenarios/05-iam-role-pattern-drift.md |
Cross-Cutting Questions
- Which runtime components should never be reimplemented by each team?
- What must be versioned and promoted like code even if it is "just configuration"?
- How will on-call know whether every service is using the same standard?
- Which controls belong in shared libraries versus IaC modules versus CI gates?
- What is the minimum platform contract new teams must adopt before they can launch?
Suggested Study Order
Read Scenario 1, Scenario 2, and Scenario 4 first if you want the strongest examples of platform drift. Use Scenario 3 to strengthen observability standards and Scenario 5 to connect standardization back to security and compliance.