normalization and denormalization Shipped

What this lens looks for

Ensures schema starts at 3NF (no transitive dependencies, no update anomalies), allows selective denormalization only for measured hotspots, documents SQLite-specific embedded tradeoff (zero network latency), and assesses denormalization interaction with table sync.

What its verifier checks

• Schema design starts from 3NF baseline; normalization violations documented if intentional
• Any denormalized data justified by measured query performance bottleneck (benchmark provided, not assumed)
• Denormalization maintenance strategy documented: triggers for automatic sync, or explicit application logic for batch updates
• Denormalized data never used for primary filtering (query logic doesn't depend on correctness of denormalized values)
• If column is denormalized, single source of truth identified (original table where it can be reconstructed)
• For synced tables: denormalization strategy compatible with sync direction and frequency
• Triggers used to maintain denormalized data are wrapped in transaction and tested
• No "just in case" denormalization; measure the query before denormalizing