Governance & Catalog

Data Exploration & Unity Catalog

Skills: databricks-unity-catalog, databricks-agent-skill-databricks MCP Tools: manage_uc_objects, get_table_details, execute_sql

List all catalogs in my workspace.

Show me all schemas in the main catalog.

List all tables in main.default and show their row counts.

Get detailed schema information for the table main.sales.transactions — show me
column names, types, and any comments.

What tables exist in main.analytics? Show me the schema for each one.

Query the system.access.audit table to show who accessed the
main.production.customers table in the last 7 days.

Query system.information_schema.tables to find the 20 largest tables by size
across all catalogs.

Show me the column-level lineage for main.gold.revenue_summary — what upstream
tables feed into it?

Query system.billing.usage to show my workspace's DBU consumption by SKU for
the past 30 days.

Permissions & Security

MCP Tools: manage_uc_grants, manage_uc_security_policies

Grant SELECT permission on main.analytics schema to the group "data-analysts".

Show me all grants on the table main.production.customers.

Set up row-level security on main.hr.employees so users can only see records
from their own department.

Create a column mask on main.finance.payroll to redact the salary column for
non-finance users.

Tags & Metadata

MCP Tools: manage_uc_tags

Add tags {"pii": "true", "data_owner": "privacy-team"} to the table
main.customers.profiles.

List all tables tagged with "pii" across the main catalog.

Volume File Operations

MCP Tools: list_volume_files, upload_to_volume, download_from_volume, create_volume_directory, delete_volume_file, delete_volume_directory

List all files in /Volumes/main/raw/incoming/ and show their sizes and
modification dates.

Upload my local file data/export.csv to
/Volumes/main/staging/uploads/export.csv.

Download the file /Volumes/main/reports/quarterly_summary.pdf to my local
machine.

Create a new directory /Volumes/main/raw/2024-01/ for organizing incoming
data files.

Iceberg Tables & Interoperability

Skills: databricks-iceberg MCP Tools: execute_sql, manage_uc_objects

Create a managed Iceberg table in main.lake.events using the iceberg table
format with partitioning on event_date.

Enable External Iceberg Reads (UniForm) on my existing Delta table
main.analytics.daily_metrics so external engines can read it as Iceberg.

Configure the Iceberg REST Catalog (IRC) so my Spark on EMR cluster can read
Unity Catalog tables as Iceberg tables.

Set up a Streaming Table with compatibility mode enabled so it can be read as
Iceberg by external engines.

Show me how to read my Databricks Delta table from Snowflake using Unity Catalog
as an Iceberg REST Catalog.

Create an Iceberg table and demonstrate reading it with PyIceberg from a local
Python environment.

MCP Tools: manage_uc_sharing

Create a share called "partner_data" and add the tables
main.analytics.product_metrics and main.analytics.daily_kpis to it.

Add a recipient "partner-corp" to my Delta Share with token-based authentication.

Data Quality Monitoring

MCP Tools: manage_uc_monitors

Create a quality monitor on main.production.orders that tracks data drift, null
rates, and column statistics on a daily schedule.

External Connections

MCP Tools: manage_uc_connections

Create a Lakehouse Federation connection to my external PostgreSQL database for
cross-database querying.