agent-bricks

Package Databricks KA, Genie Space, and MAS components into reusable Agent Bricks.

27|9|Updated Feb 3, 2026
One-click install
npx skills add https://github.com/datasciencemonkey/claude-code-cli-bricks --skill agent-bricks-datasciencemonkey
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-bricks
Source: https://github.com/datasciencemonkey/claude-code-cli-bricks/tree/main/.claude/skills/agent-bricks
Command: npx skills add https://github.com/datasciencemonkey/claude-code-cli-bricks --skill agent-bricks-datasciencemonkey

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Databricks teams often need to stitch together conversational AI capabilities across document-based Q&A, SQL exploration, and multi-agent orchestration. This Skill provides a ready-to-use blueprint—the Agent Bricks—which packages Knowledge Assistants (KA), Genie Spaces for NL-to-SQL, and Multi-Agent Supervisors (MAS) into reusable building blocks.

Core Features & Use Cases

  • KA for document Q&A over Unity Catalog volumes, with instruction templates and example workflows.
  • Genie Space guidance for turning natural language queries into SQL over Unity Catalog tables.
  • MAS orchestration that routes queries to specialized endpoints (KA, Genie, or custom models) to deliver coordinated responses.
  • MCP tooling overview: create_or_update_ka, create_or_update_genie, create_or_update_mas to provision and manage bricks in Databricks environments.

Quick Start

Prepare your data (docs in a volume and raw data for Genie), then create KA, Genie Space, and MAS bricks using the MCP tools. Wait for provisioning to ONLINE status and optionally add example questions.

Frequently Asked Questions about agent-bricks

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a Databricks conversational AI agent for document Q&A and SQL exploration?

You can build a conversational AI agent in Databricks by packaging a Knowledge Assistant for document Q&A and a Genie Space for natural language to SQL querying using pre-built agent bricks.

What is a Multi-Agent Supervisor in Databricks and how does it route queries?

A Multi-Agent Supervisor in Databricks orchestrates routing by directing user queries to specialized endpoints like Knowledge Assistants or Genie Spaces to deliver coordinated conversational AI responses.

How do I create a Knowledge Assistant from a Unity Catalog volume in Databricks?

To create a Knowledge Assistant, prepare your documents in a Unity Catalog volume and use the create_or_update_ka MCP tool to provision the agent brick for document-based Q&A.

Can I use Genie Spaces to translate natural language queries into SQL over Unity Catalog tables?

Yes, Genie Spaces enable natural language to SQL exploration, allowing you to query raw data stored in Unity Catalog tables through conversational prompts.

What is the best way to provision and manage multiple Databricks agent bricks end-to-end?

The best way to manage Databricks agent bricks is using MCP tooling interfaces like create_or_update_ka, create_or_update_genie, and create_or_update_mas to provision components until they reach ONLINE status.

Do I need to prepare data differently for Knowledge Assistants and Genie Spaces?

Yes, you must store documents in a Unity Catalog volume for Knowledge Assistants, while Genie Spaces require raw data structured in Unity Catalog tables for natural language SQL querying.