Skills that teach Claude Code how to work effectively with Databricks - providing patterns, best practices, and code examples that work with Databricks MCP tools.
Run in your project root:
# Install all skills (Databricks + MLflow)
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash
# Install specific skills
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- asset-bundles agent-evaluation
# Pin MLflow skills to a specific version
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --mlflow-version v1.0.0
# List available skills
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --listThis creates .claude/skills/ and downloads all skills. Claude Code loads them automatically.
- Databricks skills are downloaded from this repository
- MLflow skills are fetched dynamically from github.com/mlflow/skills
Manual install:
mkdir -p .claude/skills
cp -r ai-dev-kit/databricks-skills/agent-bricks .claude/skills/- agent-bricks - Knowledge Assistants, Genie Spaces, Multi-Agent Supervisors
- databricks-genie - Genie Spaces: create, curate, and query via Conversation API
- model-serving - Deploy MLflow models and AI agents to endpoints
- unstructured-pdf-generation - Generate synthetic PDFs for RAG
- vector-search - Vector similarity search for RAG and semantic search
📊 MLflow (from mlflow/skills)
- agent-evaluation - End-to-end agent evaluation workflow
- analyze-mlflow-chat-session - Debug multi-turn conversations
- analyze-mlflow-trace - Debug traces, spans, and assessments
- instrumenting-with-mlflow-tracing - Add MLflow tracing to Python/TypeScript
- mlflow-onboarding - MLflow setup guide for new users
- querying-mlflow-metrics - Aggregated metrics and time-series analysis
- retrieving-mlflow-traces - Trace search and filtering
- searching-mlflow-docs - Search MLflow documentation
- aibi-dashboards - AI/BI dashboards (with SQL validation workflow)
- databricks-unity-catalog - System tables for lineage, audit, billing
- spark-declarative-pipelines - SDP (formerly DLT) in SQL/Python
- databricks-jobs - Multi-task workflows, triggers, schedules
- synthetic-data-generation - Realistic test data with Faker
- asset-bundles - DABs for multi-environment deployments
- databricks-app-apx - Full-stack apps (FastAPI + React)
- databricks-app-python - Python web apps (Dash, Streamlit, Flask)
- databricks-python-sdk - Python SDK, Connect, CLI, REST API
- databricks-config - Profile authentication setup
- lakebase-provisioned - Managed PostgreSQL for OLTP workloads
- databricks-docs - Documentation index via llms.txt
┌────────────────────────────────────────────────┐
│ .claude/skills/ + .claude/mcp.json │
│ (Knowledge) (Actions) │
│ │
│ Skills teach HOW + MCP does it │
│ ↓ ↓ │
│ Claude Code learns patterns and executes │
└────────────────────────────────────────────────┘
Example: User says "Create a sales dashboard"
- Claude loads
aibi-dashboardsskill → learns validation workflow - Calls
get_table_details()→ gets schemas - Calls
execute_sql()→ tests queries - Calls
create_or_update_dashboard()→ deploys - Returns working dashboard URL
Create your own in .claude/skills/my-skill/SKILL.md:
---
name: my-skill
description: "What this teaches"
---
# My Skill
## When to Use
...
## Patterns
...Skills not loading? Check .claude/skills/ exists and each skill has SKILL.md
Install fails? Run bash install_skills.sh or check write permissions
- databricks-tools-core - Python library
- databricks-mcp-server - MCP server
- Databricks Docs - Official documentation
- MLflow Skills - Upstream MLflow skills repository