/bq-design-table

description: Design a BigQuery table schema with optimal partitioning, clustering, and data types. Describe your data or paste a sample.
argument-hint: "<data description or sample>"

BigQuery Table Design

Design an optimal BigQuery table schema using the `bigquery-schema-design` skill.

Instructions

1. **Determine the input:**

• If an argument is provided, treat it as the data description or sample data.
• If a file path is provided, read that file for sample data or an existing schema.
• If no argument is provided, ask the user to describe their data, its volume, and primary query patterns.

2. **Gather context** by asking (if not already provided):

• What queries will run against this table? (filter columns, aggregations, joins)
• Approximate data volume and growth rate (rows/day, total size).
• Write pattern: batch loads, streaming inserts, or frequent updates?
• Data retention requirements (how long to keep data).

3. **Design the schema** using the `bigquery-schema-design` skill:

• Choose partitioning strategy based on primary filter column and data volume.
• Choose clustering columns based on secondary filters and join keys.
• Decide nested vs. flat based on entity relationships and update patterns.
• Select optimal data types for each column.
• Consider table type (native, materialized view, external) based on access patterns.

4. **Output the recommendation** in the standard schema design format:

• Design decisions summary (partitioning, clustering, nesting, data types).
• Complete `CREATE TABLE` DDL with all options.
• Rationale explaining trade-offs and why this design fits the workload.
• If applicable, suggest materialized views or complementary tables.