Data Warehouse Tools

The Data Warehouse Tools allow DinoAI to explore your connected warehouse metadata directly from the Code IDE. DinoAI can inspect databases, schemas, tables, columns, and relationships so it can write more accurate SQL, generate dbt assets faster, and reason about your warehouse structure with real context.

Metadata only. These tools inspect schema metadata, not table contents. If you need DinoAI to run queries against actual data, use the SQL Execution Tool.

Capabilities

The Data Warehouse Tools give DinoAI the ability to:

• Discover available databases, catalogs, schemas, and tables

• Inspect column names, data types, descriptions, and nested fields where supported

• Understand relationships between warehouse objects

• Generate accurate dbt source definitions and model references

• Suggest join conditions based on real warehouse structure

• Help investigate warehouse-specific metadata such as query performance where supported

Supported Data Warehouses

Each warehouse has its own toolset and capabilities:

• BigQuery Tools

• ClickHouse Tools

• Databricks Tools

• Microsoft Fabric Tools

• Redshift Tools

• Snowflake Tools

• Snowflake Query Analysis Tool

• SQL Server Tools

• Trino Tools

Requires a warehouse connection. These tools are available when your workspace has a supported warehouse configured. See Code IDE environment to set up a connection.

Using the Data Warehouse Tools

1. Open DinoAI in the right panel of the Code IDE

2. Describe the warehouse object you want to inspect

3. Add your prompt describing what DinoAI should do with that metadata

4. Grant permission when DinoAI asks to access your warehouse connection

5. Review the results and apply DinoAI's suggested changes

Example Use Cases

Generating dbt Source Definitions

Prompt

Create a sources.yml file for all tables in the marketing schema.

Result: DinoAI inspects the schema metadata and generates source definitions with the correct table and column names.

Understanding an Unfamiliar Schema

Prompt

List the tables in the analytics schema and explain how they relate.

Result: DinoAI explores the warehouse structure, summarizes the main tables, and highlights likely joins and dependencies.

Building a Model with Accurate References

Prompt

Create a staging model for the customers table with standardized column names.

Result: DinoAI uses the real column definitions to create a model with accurate references and cleaner naming.

Working with Other Tools

The Data Warehouse Tools work well alongside DinoAI's other capabilities:

• Combine with the SQL Execution Tool to validate assumptions against live query results

• Combine with the Column Level Lineage tool to trace how warehouse columns flow through downstream dbt assets

• Use with .dinorules to define persistent modeling and naming conventions for DinoAI

Best Practices

• Be specific about schemas when working in large warehouses

• Provide full object names when possible to reduce extra lookups

• Mention the warehouse platform if your workspace uses multiple connections

• Start with exploration before asking DinoAI to generate models or sources

• Use the warehouse-specific pages above for features unique to each platform