Kubernetes Cluster Manager

Decision Trees and Conditional Logic

• If compliance or regulatory scope applies -> add required controls and audit steps.
• If latency budget is strict -> choose low-latency storage and caching.
• Else -> prefer cost-optimized storage and tiering.
• If data consistency is critical -> prefer transactional boundaries and strong consistency.
• Else -> evaluate eventual consistency or async processing.

Error Handling and Edge Cases

• Partial failures across dependencies -> isolate blast radius and retry with backoff.
• Data corruption or loss risk -> enable backups and verify restore path.
• Limited access to systems -> document gaps and request access early.
• Legacy dependencies with limited change tolerance -> use adapters and phased rollout.

Tool Requirements and Dependencies

• CLI and SDK tooling for the target stack
• Credentials or access tokens for required environments
• Diagramming or spec tooling when producing docs

Stack Profiles

• Use Profile A, B, or C from skills/STACK_PROFILES.md.
• Note selected profile in outputs for traceability.

Validation

• Requirements coverage check
• Security and compliance review
• Performance and reliability review
• Peer or stakeholder sign-off

Rollback Procedures

• Revert config or deployment to last known good state.
• Roll back database migrations if applicable.
• Verify service health, data integrity, and error rates after rollback.

Success Metrics

• Measurable outcomes (latency, error rate, uptime, cost)
• Acceptance thresholds defined with stakeholders

Example Workflows and Use Cases

• Minimal: apply the skill to a small service or single module.
• Production: apply the skill to a multi-service or multi-tenant system.