Data Warehousing
When to use
Use this skill for data warehousing when you need to organize analytical models by grain and business meaning. The expected result is a warehouse model and metric definitions.
Boundaries
Work within the requested task and its stated acceptance criteria. Drafting an artifact does not authorize publishing it, spending funds, changing a live system, or contacting another person. Identify any such action separately before taking it.
Inputs
Inspect schemas, data lineage, row grain, volumes, freshness, and downstream consumers. Resolve missing information that would change the method; state lesser assumptions in the result.
Method
- Diagnose. Map tables or collections, keys, row grain, indexes, volumes, consumers, and expected freshness.
- Decide. Choose a transformation or storage design that preserves contracts and can be replayed.
- Produce. Build a warehouse model and metric definitions from the inspected material; keep assumptions distinguishable from observed facts.
Decision rules
- Specify nulls, duplicates, late data, schema drift, and replay behavior before transformation or migration.
- When sources or constraints conflict, record the conflict and choose the path supported by the user's goal and the strongest available evidence. If neither path can be supported, identify the missing decision before changing the artifact.
Domain rules
- Define null, duplicate, late-arrival, and deletion behavior.
- Document migration, backfill, and rollback expectations.
Verification
Reconcile aggregates to trusted source totals. Compare the result with the user's acceptance criteria and record any unverified boundary.
Reconcile sample and aggregate results; include query plan or pipeline metrics and rollback or backfill steps.
Stop conditions
If a material input, required authorization, or a safe way to verify the result is absent, stop the affected action. Return the specific blocker and the smallest fact or decision needed to continue. Do not report an unrun check as passed.
Output
Provide a warehouse model and metric definitions. Include the decisive evidence and actual verification result. Name any artifact location and unresolved issue that affects its use.
