This toolkit enhances AI-driven development on the Databricks platform, providing AI coding assistants like Claude Code and Cursor with access to trusted data sources and Databricks functionalities. It aims to accelerate the development of complex Databricks applications, from data pipelines to AI dashboards and full-stack web apps, by integrating AI agents directly into the development workflow.

How It Works

The AI Dev Kit offers a multi-component approach: a core Python library (databricks-tools-core) for programmatic access, an MCP server exposing over 50 Databricks tools for AI assistants, markdown-based "skills" for guiding AI agents on Databricks patterns, and a full-stack web application (databricks-builder-app) with a chat UI for Databricks development. This architecture allows AI agents to leverage Databricks capabilities securely and efficiently, acting as a bridge between AI models and the Databricks ecosystem.

Quick Start & Requirements

• Primary Install: Installation is handled via shell scripts. For Mac/Linux: bash <(curl -sL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/install.sh). For Windows (PowerShell): irm https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/install.ps1 | iex. Advanced options allow global installation, profile specification, or installing specific tools. The Visual Builder App requires cd ai-dev-kit/databricks-builder-app && ./scripts/setup.sh.
• Prerequisites: uv (Python package manager), Databricks CLI, and an AI coding environment (e.g., Claude Code, Cursor).
• Links:
  • Mac/Linux Install Script: https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/install.sh
  • Windows Install Script: https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/install.ps1

Highlighted Details

• Enables AI-driven development for building Spark Declarative Pipelines, Databricks Jobs, AI/BI Dashboards, Unity Catalog assets, Genie Spaces, Knowledge Assistants (RAG), MLflow Experiments, Model Serving, and Databricks Apps.
• Provides a core Python library (databricks-tools-core) for direct integration into Python projects, compatible with frameworks like LangChain and OpenAI Agents SDK.
• Includes a databricks-mcp-server exposing over 50 tools for AI assistants and databricks-skills offering Databricks best practices.
• Features a full-stack web application (databricks-builder-app) with a chat UI for enhanced Databricks development experiences.

Maintenance & Community

This project is provided by Databricks Field Engineering. No specific community channels (e.g., Discord, Slack) or roadmap links are detailed in the README.

Licensing & Compatibility

The primary license is the "Databricks License" (© 2026 Databricks, Inc.). However, several third-party dependencies carry their own licenses: pymupdf is AGPL-3.0, and psycopg2-binary is LGPL-3.0. These licenses, particularly AGPL-3.0, may impose copyleft restrictions on derivative works, potentially affecting compatibility with closed-source commercial applications.

Limitations & Caveats

The AGPL-3.0 license associated with pymupdf requires that any modifications or derivative works be made available under the same license, which could be a significant consideration for commercial closed-source projects. Additionally, AI coding environments like Cursor and Copilot require manual settings updates post-installation. The project's copyright year is listed as 2026.