AI coding assistant skills for building PWAF-compliant Databricks partner (ISV) integrations, following the Partner Well-Architected Framework (PWAF).
The Databricks Partner AI Dev-Kit gives your AI coding assistant (Claude Code, Cursor, etc.) the patterns and best practices needed to build secure, compliant Databricks integrations. These skills cover authentication, telemetry, Unity Catalog, and driver patterns across 10+ connectors.
- PWAF-compliant connectors with proper authentication (PAT, OAuth M2M, U2M)
- Python integrations using SDK, SQL Connector, SQLAlchemy, or Databricks Connect
- Java integrations using SDK or JDBC driver
- Go integrations using SDK or SQL Driver
- Node.js integrations using SQL Driver
- REST API clients in any language
- Telemetry-enabled partners with User-Agent attribution
- U2M browser flows with custom OAuth apps and PKCE
- AI coding environment: Claude Code or Cursor
- Git (for cloning)
Mac / Linux:
# Clone the repository
git clone https://github.com/databricks-solutions/partner-ai-dev-kit.git databricks-partner-ai-dev-kit
# Run the installer
cd databricks-partner-ai-dev-kit
bash install.shOptions:
# Install globally (all projects)
bash install.sh --global
# Install for specific tool only
bash install.sh --tool cursor
bash install.sh --tool claude
# Force reinstall (overwrite existing)
bash install.sh --forceCopy the skills/ folder to your AI assistant's skills directory:
- Claude Code:
~/.claude/skills/databricks-partner-ai-dev-kit/ - Cursor:
.cursor/rules/databricks-partner-ai-dev-kit/(project) or~/.cursor/rules/databricks-partner-ai-dev-kit/(global)
| Use Case | Skill | Category |
|---|---|---|
| Add Databricks to existing project | adding-databricks-connector | General |
| Structure a new connector | connector-structure | General |
| Python SDK (workspace APIs, UC, Jobs) | python-sdk | Python |
| Python SQL Connector (SQL warehouse) | python-sql-connector | Python |
| Python SQLAlchemy (ORM) | python-sqlalchemy | Python |
| Databricks Connect (Spark) | databricks-connect | Python |
| Java SDK (workspace APIs) | java-sdk | Java |
| Java JDBC (SQL warehouse) | java-jdbc | Java |
| Go SDK (workspace APIs) | go-sdk | Go |
| Go SQL Driver (SQL warehouse) | go-sql-driver | Go |
| Node.js SQL Driver | nodejs-sql-driver | Node.js |
| REST API (any language) | rest-api | General |
| U2M authentication flows | u2m | Auth |
| Test runner (all auth types) | connector-testing | Testing |
| Build report after implementation | build-report | Documentation |
| Component | Description |
|---|---|
| skills/ | 17+ task-focused skills + auth patterns (SKILL.md + authentication.md per connector) |
| prompts/ | Ready-to-use build prompts — one per connector |
| examples/ | 10 complete working connector implementations |
| SKILL.md | Main entry skill with overview and quick reference |
| manifest.json | Skill manifest with categories for discovery |
| Language | Connectors |
|---|---|
| Python | SDK, SQL Connector, SQLAlchemy, Databricks Connect |
| Java | SDK, JDBC |
| Go | SDK, SQL Driver |
| Node.js | SQL Driver |
| Any | REST API, ODBC |
All skills support these authentication methods per PWAF requirements:
| Auth Type | Use Case |
|---|---|
| PAT | Personal Access Token (simple, user-specific) |
| OAuth M2M | Machine-to-machine (service principals, backend) |
| OAuth U2M | User-to-machine (interactive, browser-based) |
| Token Pass-through | Pre-obtained tokens (CI/CD, headless) |
The examples/ directory contains complete, working connector implementations for
all 10 supported stacks. Each example includes code for all auth types,
.env.template, and a build_report.md.
Examples are not part of the AI build flow — they serve as human reference, validation baseline, and debugging aid. The skills are self-sufficient.
See examples/README.md for setup and run instructions.
- INDEX.md – One-line index of all rules, skills, prompts, and examples
- SKILL.md – Full reference with code examples
- examples/README.md – Working connector implementations
- PWAF Documentation
- PWAF Integration Requirements
(c) 2026 Databricks, Inc. All rights reserved.
The source in this project is provided subject to the Databricks License. See LICENSE.md for details.
See CONTRIBUTING.md for guidelines.