Data Architecture Skill

Purpose

Design data architectures including data models, pipeline designs, governance frameworks, and quality management for operational and analytical systems.

---

Parameters

Parameter	Type	Required	Validation	Default
data_domain	string	✅	min: 20 chars	-
design_type	enum	⚪	model|pipeline|governance|quality	model
data_type	enum	⚪	operational|analytical|streaming	operational
volume_tier	enum	⚪	small|medium|large|massive	medium
output_format	enum	⚪	erd|yaml|json	erd

---

Execution Flow

┌──────────────────────────────────────────────────────────┐
│ 1. VALIDATE: Check data domain and requirements          │
│ 2. DISCOVER: Identify data sources and entities          │
│ 3. MODEL: Create conceptual/logical/physical model       │
│ 4. DESIGN: Pipeline or governance framework              │
│ 5. QUALITY: Define data quality rules                    │
│ 6. VALIDATE: Check model consistency                     │
│ 7. DOCUMENT: Return data architecture                    │
└──────────────────────────────────────────────────────────┘

---

Retry Logic

Error	Retry	Backoff	Max Attempts
VALIDATION_ERROR	No	-	1
MODEL_GENERATION_ERROR	Yes	1s	2
FORMAT_ERROR	Yes	500ms	3

---

Logging & Observability

log_points:
  - event: design_started
    level: info
    data: [design_type, data_type]
  - event: entities_identified
    level: info
    data: [entity_count, relationship_count]
  - event: quality_rules_defined
    level: info
    data: [rule_count, dimensions_covered]

metrics:
  - name: models_created
    type: counter
    labels: [design_type]
  - name: design_time_ms
    type: histogram
  - name: entity_count
    type: gauge

---

Error Handling

Error Code	Description	Recovery
E401	Missing data domain	Request domain description
E402	Invalid relationships	Highlight circular/missing refs
E403	Schema validation failed	Show validation errors
E404	Unsupported volume tier	Suggest architectural changes

---

Unit Test Template

test_cases:
  - name: "E-commerce data model"
    input:
      data_domain: "E-commerce order management"
      design_type: "model"
      output_format: "erd"
    expected:
      has_entities: true
      entities_include: ["Customer", "Order", "Product"]
      has_relationships: true
      valid_erd: true

  - name: "Analytics pipeline"
    input:
      data_domain: "Customer analytics"
      design_type: "pipeline"
      data_type: "analytical"
    expected:
      has_ingestion: true
      has_transformation: true
      has_serving: true

  - name: "Data quality rules"
    input:
      data_domain: "User profiles"
      design_type: "quality"
    expected:
      has_dimensions: true
      dimensions_include: ["completeness", "accuracy"]
      has_rules: true

---

Troubleshooting

Common Issues

Symptom	Root Cause	Resolution
Missing relationships	Incomplete domain	Add missing entities
Invalid ERD syntax	Format error	Validate Mermaid ERD
Missing quality rules	Dimensions not specified	Add quality dimensions

Debug Checklist

□ Is data domain clearly defined?
□ Are all entities identified?
□ Are relationships correctly typed?
□ Is output format valid?
□ Are quality dimensions covered?

---

Data Quality Dimensions

Dimension	Example Rule
Completeness	NOT NULL checks
Accuracy	Regex validation
Consistency	Referential integrity
Timeliness	SLA monitoring
Uniqueness	Primary key constraints

---

Integration

Component	Trigger	Data Flow
Agent 06	Design request	Receives domain, returns model
Agent 04	Cloud data services	Cloud data platform

---

Quality Standards

• Normalized: 3NF for operational, denormalized for analytical
• Documented: All entities and relationships described
• Quality-first: DQ rules for all critical fields

---

Version History

Version	Date	Changes
2.0.0	2025-01	Production-grade: ERD, pipelines, DQ framework
1.0.0	2024-12	Initial release