lab-architecture-design

name	lab-architecture-design
description	Procedures for producing architecture summaries, Mermaid diagrams, module breakdowns, and file trees from extracted metadata.

Architecture Design

Procedures for designing Azure lab architectures from extracted metadata. Used by the Lab-Designer agent (Phase 2).

When to Use

• Designing a lab architecture after metadata extraction
• Creating Mermaid diagrams for lab documentation
• Planning module structure for IaC implementations
• Generating file trees for lab folders

---

R-110: Architecture Summary Procedure

Given metadata (exam, domain, topic, key_services):

1. Identify the primary learning objective from the exam question.
2. Determine which Azure services demonstrate the concept.
3. Map service dependencies (e.g., VM requires VNet → Subnet → NSG).
4. Write 2–4 sentences covering:
   • What services are deployed and why
   • How they connect/interact
   • What the lab participant will observe or validate

---

R-111: Mermaid Diagram Procedure

Apply criteria from lab-shared-contract R-013.

A Mermaid diagram is always required (lab-shared-contract R-013).

When 2+ interconnected resources exist

1. Choose layout: graph TD (top-down) for hierarchical, graph LR (left-right) for pipeline/flow.
2. Use governance-compliant resource names (lab-shared-contract R-001 / R-002 / R-003).
3. Show resource group as the root container.
4. Draw lines for dependencies (network containment, access paths, data flow).
5. Use subgraphs for logical groupings (e.g., subnets within a VNet).

When fewer than 2 interconnected resources

1. Diagram the overall process reflective of the exam question.
2. Show the end-to-end flow: data source → Azure service → output / consumer.
3. Include access methods, decision points, or output formats when they are the focus of the question.
4. Prefer graph LR (left-right) for process / pipeline flows.

Common topology patterns:

• Hub-spoke: Subgraphs for hub and spoke VNets.
• Multi-tier: Top-down flow from frontend to backend.
• AI pipeline: Left-right flow from data source through processing to output.
• Process / access flow: Left-right flow showing how users or clients interact with a single service.

---

R-112: Module Breakdown Procedure

Apply lab-shared-contract R-022 (module rule).

1. List all Azure resources from the architecture.
2. Group by domain concern:
   • Networking: VNet, Subnet, NSG, Route Table, Peering
   • Compute: VM, NIC, Disk, Availability Set
   • Storage: Storage Account, Containers, RBAC
   • AI: Cognitive Services, OpenAI, Deployments
   • Monitoring: Log Analytics, App Insights, Diagnostic Settings
3. Create one module per domain group (only if 2+ resource types in that group).
4. Define inputs: resource group name, location, tags, cross-module references.
5. Define outputs: resource IDs, endpoints, principal IDs (for RBAC wiring).

---

R-113: File Tree Procedure

Apply lab-shared-contract R-010 (lab folder structure).

1. Select the matching structure template for the deployment method.
2. Replace <EXAM>, <domain>, <topic> with actual values (lowercase).
3. Add module subfolders matching the module breakdown from R-112.
4. Add validation script file(s).
5. Verify the tree is complete: README, all code files, validation.

---

R-114: When-to-Use Criteria

Use this skill when:

• Metadata extraction is complete and architecture design is needed
• A Mermaid diagram needs to be generated
• Module structure needs to be planned
• A file tree needs to be generated for a new lab