AI agent tools

AI agent tools give your agents practical capabilities like searching through documents, analyzing data in spreadsheets, calling REST APIs, or executing Python code. Instead of just generating text, your agent can work with real data and systems.

Choose your tool approach

You have three options for creating tools that your agents can use. Choose based on your needs for governance, flexibility, or integration:

Unity Catalog function tool	Agent code tool	Model Context Protocol (MCP) tool
Defined as Unity Catalog UDFs Managed in Unity Catalog as a central registry for tools Built-in security and compliance features Grants easier discoverability and reuse Good for applying transformations and aggregations on large datasets See Create and use UC tools	Defined directly in your agent's code Good for calling REST APIs, running arbitrary code, or running low-latency tools Lacks the built-in governance and discoverability of Unity Catalog functions See Example agent code tool	Tools that follow the MCP standard for tool interoperability Can be managed (Databricks-hosted), external (third-party), or custom (self-hosted) Standardized interface allows reuse across different agent frameworks Good for accessing external services, third-party APIs, or building reusable tool ecosystems See MCP on Databricks

You can mix and match any of these approaches. All three work whether you're building agents in plain Python or using frameworks like LangChain, OpenAI SDK, or LangGraph.

Common tool patterns

Explore these tool patterns and examples, which can be implemented using any of the three tool approaches:

Tool pattern	Description
Structured data retrieval tools	Query SQL tables, databases, and structured data sources.
Unstructured data retrieval tools	Search document collections and perform retrieval augmented generation.
Code interpreter tools	Allow agents to run Python code for calculations, data analysis, and dynamic processing.
External connection tools	Connect to external services and APIs such as Slack.
AI Playground prototyping	Use the AI Playground to quickly add Unity Catalog tools to agents and prototype their behavior before deploying them.