Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

README.md

Databricks Skills for Claude Code

Skills that teach Claude Code how to work effectively with Databricks - providing patterns, best practices, and code examples that work with Databricks MCP tools.

Installation

Run from your project root (the directory where you want .claude/skills created).

From this repository (local script)

If you already have the repo (fork or clone), use the script on disk:

# Install all skills (Databricks + MLflow + APX) — downloads from GitHub by default
./databricks-skills/install_skills.sh

# Install Databricks skills only from this checkout (no network for those skills)
./databricks-skills/install_skills.sh --local

# Install specific skills
./databricks-skills/install_skills.sh databricks-bundles agent-evaluation

# Pin MLflow / APX versions
./databricks-skills/install_skills.sh --mlflow-version v1.0.0

# List available skills
./databricks-skills/install_skills.sh --list

# Install + upload to workspace for Genie Code (/Workspace/Users/<you>/.assistant/skills)
./databricks-skills/install_skills.sh --install-to-genie

./databricks-skills/install_skills.sh --install-to-genie --profile prod

# Local Databricks skills + Genie upload
./databricks-skills/install_skills.sh --local --install-to-genie

Paths assume you are at the ai-dev-kit repo root. From another project, copy or symlink the script, or use the curl flow below.

Without cloning (curl)

Use this when you only want the installer and not the full repo:

# Install all skills
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash

# Install specific skills (pass args after bash -s --)
curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- databricks-bundles agent-evaluation

curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --mlflow-version v1.0.0

curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --list

curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --install-to-genie

curl -sSL https://raw.githubusercontent.com/databricks-solutions/ai-dev-kit/main/databricks-skills/install_skills.sh | bash -s -- --install-to-genie --profile prod

--install-to-genie uploads the tree under ./.claude/skills to the workspace (requires the databricks CLI).

This 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/databricks-agent-bricks .claude/skills/

Available Skills

🤖 AI & Agents

  • databricks-ai-functions - Built-in AI Functions (ai_classify, ai_extract, ai_summarize, ai_query, ai_forecast, ai_parse_document, and more) with SQL and PySpark patterns, function selection guidance, document processing pipelines, and custom RAG (parse → chunk → index → query)
  • databricks-agent-bricks - Knowledge Assistants, Genie Spaces, Supervisor Agents
  • databricks-genie - Genie Spaces: create, curate, and query via Conversation API
  • databricks-model-serving - Deploy MLflow models and AI agents to endpoints
  • databricks-unstructured-pdf-generation - Generate synthetic PDFs for RAG
  • databricks-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

📊 Analytics & Dashboards

  • databricks-aibi-dashboards - Databricks AI/BI dashboards (with SQL validation workflow)
  • databricks-unity-catalog - System tables for lineage, audit, billing

🔧 Data Engineering

  • databricks-iceberg - Apache Iceberg tables (Managed/Foreign), UniForm, Iceberg REST Catalog, Iceberg Clients Interoperability
  • databricks-spark-declarative-pipelines - SDP (formerly DLT) in SQL/Python
  • databricks-jobs - Multi-task workflows, triggers, schedules
  • databricks-synthetic-data-gen - Realistic test data with Faker

🚀 Development & Deployment

  • databricks-bundles - DABs for multi-environment deployments
  • databricks-app-apx - Full-stack apps (FastAPI + React)
  • databricks-app-python - Python web apps (Dash, Streamlit, Flask) with foundation model integration
  • databricks-python-sdk - Python SDK, Connect, CLI, REST API
  • databricks-config - Profile authentication setup
  • databricks-lakebase-provisioned - Managed PostgreSQL for OLTP workloads

📚 Reference

  • databricks-docs - Documentation index via llms.txt

How It Works

┌────────────────────────────────────────────────┐
│  .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"

  1. Claude loads databricks-aibi-dashboards skill → learns validation workflow
  2. Calls get_table_stats_and_schema() → gets schemas
  3. Calls execute_sql() → tests queries
  4. Calls manage_dashboard(action="create_or_update") → deploys
  5. Returns working dashboard URL

Custom Skills

Create your own in .claude/skills/my-skill/SKILL.md:

---
name: my-skill
description: "What this teaches"
---

# My Skill

## When to Use
...

## Patterns
...

Troubleshooting

Skills not loading? Check .claude/skills/ exists and each skill has SKILL.md

Install fails? Run bash install_skills.sh or check write permissions

Related