AI Plugin Indexdata platform
Databricks MCP servers and skills
There are 20 ways to connect Databricks to Claude, ChatGPT, Cursor and other AI assistants: 7 MCP servers and 13 skills. The most used is AI Sdk Provider (115 GitHub stars).
Best Databricks MCP servers, compared →Most used
Databricks provider for Vercel AI SDK
How the dispatcher writes a one-line handoff and merges specialist replies. Per-agent skill — always visible to the query agent without opting in.
Create Databricks resources that agents connect to as tools. Use when: (1) User needs to create a Genie space, vector search index, UC function, or UC connection, (2) User says 'create tool', 'set up genie', 'create vector search', 'register MCP server', (3) Before add-tools when the resource doesn't exist yet, (4) User asks 'what do I need to create before adding this tool'.
Standalone MCP (Model Context Protocol) server for Databricks SQL using @databricks/sql
Model Context Protocol (MCP) server for interacting with Databricks
Use this skill to act as the agent defined in databricks-notebook-publisher.md for tasks matching that role.
MCP servers (7)
- 1AI Sdk Provider databricks · Cloud and DevOpsDatabricks provider for Vercel AI SDK76
- 2Databricks MCP victoprincipe · DatabasesStandalone MCP (Model Context Protocol) server for Databricks SQL using @databricks/sql56
- 3Databricks MCP Server characat0 · Knowledge and memoryModel Context Protocol (MCP) server for interacting with Databricks54
- 4Databricks MCP Server Linux Amd64 characat0Platform-specific binary for databricks-mcp-server49
- 5Databricks MCP Server pramodbhatofficial · Databases263 tools for Databricks: Unity Catalog, SQL, Compute, Jobs, Serving, and more.30
- 6Databricks Field Guide by TechFabric databricks · Search and webSearch and read TechFabric's Databricks Field Guide: chapters, code examples and what changed.16
- 7Databricks mcp-dir · DatabasesYour Databricks Lakehouse in natural language: run SQL on your SQL warehouses, track long-running qu12
Skills (13)
- 1routing-brief databricks · AI and LLM toolsHow the dispatcher writes a one-line handoff and merges specialist replies. Per-agent skill — always visible to the query agent without opting in.69
- 2create-tools databricks · Search and webCreate Databricks resources that agents connect to as tools. Use when: (1) User needs to create a Genie space, vector search index, UC function, or UC connection, (2) User says 'create tool', 'set up genie', 'create vector search', 'register MCP server', (3) Before add-tools when the resource doesn't exist yet, (4) User asks 'what do I need to create before adding this tool'.65
- 3databricks-notebook-publisher alexeyban · Data and analyticsUse this skill to act as the agent defined in databricks-notebook-publisher.md for tasks matching that role.50
- 4databricks-ai-runtime jotap-rocha · Cloud and DevOpsDatabricks AI Runtime (`air`) CLI — the command-line tool for submitting and managing GPU training workloads on Databricks serverless compute. Use for: running `air` workloads, custom Docker image setup, environment configuration, and troubleshooting `air` jobs.33
- 5discover-tools mkgs-databricks-demos · Search and webDiscover available tools and resources in Databricks workspace. Use when: (1) User asks 'what tools are available', (2) Before writing agent code, (3) Looking for MCP servers, Genie spaces, UC functions, or vector search indexes, (4) User says 'discover', 'find resources', or 'what can I connect to'.30
- 6databricks-config leary-pokenManage Databricks workspace connections: check which workspace you're connected to, switch workspaces, list available workspaces, or authenticate to a new workspace.28
- 7databricks-file-management kendrick · Documents and filesFile upload, download, and directory management via the modern Files API or legacy DBFS.27
- 8eventhub-analyze-databricks-streaming javakishore-veleti · Cloud and DevOpsAnalyze Azure Databricks or Spark Structured Streaming integrations with Azure Event Hubs, including Kafka-compatible endpoints, offsets, checkpoints, partition parallelism, lag, throughput, and downstream processing. Use for Databricks/Spark streaming pipeline issues.26
- 9databricks-app-production-readiness InsightpulseaiDatabricks App Production Readiness skill22
- 10databricks-production-readiness-judge InsightpulseaiDatabricks Production Readiness Judge skill22
- 11dataeng-databricks-ingestion juankmvanegas · Cloud and DevOpsImplement Azure Databricks ingestion patterns for cloud storage, JDBC, Kafka, Lakeflow Connect, Auto Loader, CDC, streaming, and external systems with secure credentials and schema drift handling.22
- 12databricks 11vated · ProductivityDatabricks - use when working on related tasks19
- 13data-addon-databricks-mcp Camilool8 · SecurityDatabricks Managed MCP conventions (Public Preview) — Unity Catalog ACLs as the upstream gate, OAuth only never a personal access token, Genie spaces treated as downstream lineage consumers, and Mosaic AI model serving published by humans/CI not the agent. Use when querying Databricks, wiring the Databricks MCP, or working with Unity Catalog.19
Questions
Is there an official Databricks MCP server?
Check the publisher on each listing: servers published by Databricks itself are the official ones. The most used Databricks server here is AI Sdk Provider by databricks.
How do I connect Databricks to Claude?
Pick a server below, open its page and copy the install command for Claude Code, or the config block for Claude Desktop. Most Databricks servers need an API key or token from your Databricks account.