Knowledge Base
The Knowledge Base tab houses all of the metadata that the Querio agent uses to work with your data. It is divided into three main parts:
Data Catalog
The Data Catalog shows all the tables in your database along with their corresponding columns and some key pieces of metadata.
Golden Queries
Golden Queries are SQL or Python snippets that will explain to Querio how to work with your data and make more specific calculations.
Macros
Macros are shortcuts that let you quickly reference specific values from your database columns when writing prompts in the Explore page.
Data Catalog
The Data Catalog shows all the tables in your database along with their corresponding columns and some key pieces of metadata.
Explore Visibility
Every table and column on Querio has an AI agent visibility setting that can be set to true or false. True means that the AI agent in the Explore page can see and interact with this table or column using SQL. You can see and change this setting using these buttons next to tables and columns.

Descriptions and Other Metadata
The database as a whole, tables and columns all have descriptions and other metadata you can update to explain to Querio how to work with your data. Admins can edit that metadata by clicking any of the edit icons.

Bulk Editing
You might want to make many updates at once and to make that faster you can use the JSON view. This will let you make updates to all table and column metadata in a single go and save them in bulk. Simply toggle JSON view, make your updates and save them.

Adding New Tables
If you need to add a new table to the data catalog, follow these steps:
- Click the the Add new tables button

- Select the tables you want to generate metadata for

- Review and make manual updates to the metadata before adding this table to your data catalog

Golden Queries
Golden Queries are SQL or Python snippets that will explain to Querio how to work with your data and make more specific calculations.

Macros
Macros are shortcuts that let you quickly reference specific values from your database columns when writing prompts in the Explore page.
