discover-tools

Discover Databricks tools, UC functions, vector search indexes, Genie spaces, and MCP servers.

2|Updated Apr 6, 2026
One-click install
npx skills add https://github.com/bcheng004/agent-langgraph-advanced-workshop --skill discover-tools-bcheng004
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: discover-tools
Source: https://github.com/bcheng004/agent-langgraph-advanced-workshop/tree/main/.claude/skills/discover-tools
Command: npx skills add https://github.com/bcheng004/agent-langgraph-advanced-workshop --skill discover-tools-bcheng004

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates time wasted guessing what Databricks resources or integrations are available before you build an agent, so you can connect the right tools confidently.

Core Features & Use Cases

  • UC Functions as Agent Tools: Discover SQL UDFs exposed via MCP endpoints so the agent can call them directly.
  • Vector Search Indexes for RAG: Identify vector search indexes to wire retrieval into your agent workflow.
  • Genie Spaces and MCP Servers: Find Genie spaces and both custom/external MCP servers to expand capability beyond UC.
  • Use Case: When you need a retrieval + database workflow, discover available UC functions and vector search indexes first, then implement the agent tools that match your environment.

Quick Start

Ask the agent to run tool discovery for your current workspace so it returns the available UC functions, vector search indexes, Genie spaces, and MCP servers to connect.

Frequently Asked Questions about discover-tools

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I discover available Databricks tools and integration endpoints for agent development?

You can discover available Databricks tools by running the discover-tools command with optional catalog and schema scoping to locate UC functions, vector search indexes, Genie spaces, and MCP servers for agent connectivity. The command optionally emits JSON or a markdown report for easy integration.

What is the best way to find vector search indexes and UC functions before building a RAG agent?

Finding vector search indexes and UC functions for RAG agents requires running a tool discovery process against your Databricks workspace. This identifies available SQL UDFs exposed via MCP endpoints and vector search indexes so you can wire retrieval into your agent workflow.

Can I search for Genie spaces and MCP servers in a specific Databricks catalog?

Yes, you can search for Genie spaces and MCP servers within a specific Databricks catalog by running the discover-tools command with optional catalog and schema scoping parameters. This filters the discovery results to match your environment and returns the correct MCP URL pattern outputs.

How do I list SQL UDFs exposed via MCP endpoints for direct agent integration?

To list SQL UDFs exposed via MCP endpoints, run tool discovery for your current workspace. The process returns the available UC functions that agents can call directly, eliminating the need to guess what Databricks resources or integrations are available.

Does tool discovery support exporting results as JSON for programmatic agent integration?

Yes, tool discovery supports emitting JSON or a markdown report as output formats. This allows you to programmatically parse the discovered UC functions, vector search indexes, Genie spaces, and MCP servers when wiring capabilities into your agent code.

What Databricks resources should I locate before writing agent code for a retrieval and database workflow?

Before writing agent code for a retrieval and database workflow, you should locate available UC functions and vector search indexes using tool discovery. This ensures you connect the correct Databricks tools and integration endpoints confidently without guessing available resources.