AI-Driven Catalog
MetaBot AI features, the DataGalaxy MCP Server, and other AI-powered capabilities.
25 articles
Blink, a data governance chatbot
Functional boundaries and limitations of Blink
A quick reference to what Blink can and can't do operationally — scope, permissions, SQL generation, and how to work around its limits.
Best Prompting Practices for Blink
A practical guide to writing prompts that get the best results from Blink, DataGalaxy's data governance chatbot.
Technical Overview: Blink in the DataGalaxy Platform
How Blink fits into DataGalaxy's secure, multi-tenant, Kubernetes-based SaaS architecture, and the models behind it.
Why Chatbot is Called Blink
Why DataGalaxy's data governance chatbot is named Blink, and what that name says about its purpose.
Blink, DataGalaxy Chatbot helping to deal with data governance
Meet Blink, DataGalaxy's in-app AI chatbot that helps stewards and readers explore data and get quick answers.
MCP Server
How to Install the DataGalaxy MCP Server as a Claude Extension
Install the DataGalaxy MCP Server in Claude Desktop as a one-click .mcpb Desktop Extension — no config file editing required.
Adding DataGalaxy MCP Server to Mistral Vibe
Step-by-step guide to adding the DataGalaxy MCP Server as a custom connector in Mistral Vibe (formerly Le Chat), with OAuth 2.1 (recommended) or Personal Access Token authentication.
Metadata+Data: Using DataGalaxy MCP Server with your Data Storage MCP (Databricks Example)
Combine DataGalaxy's metadata MCP with a data-storage MCP like Databricks in Cursor: find a documented query in the catalog, then run it against live data.
How to Connect the DataGalaxy MCP Server to Microsoft Copilot (Copilot Studio) and Publish to Microsoft Teams
Full walkthrough for connecting the DataGalaxy MCP Server to a Microsoft Copilot Studio agent and publishing it to Teams and Microsoft 365 Copilot.
MCP Client Prompt Template
A ready-to-use system prompt template for any MCP client, updated to match the current DataGalaxy MCP tool set.
Prerequisites: Microsoft Copilot Studio licensing for MCP Server
The Microsoft licensing requirements — maker, tenant, and end-user — needed to connect the DataGalaxy MCP Server to a Copilot Studio agent.
MCP Server Instructions
The system-prompt-style instruction block DataGalaxy uses to steer MCP clients toward the right tool, current with the live tool set.
Use Cases: How the DataGalaxy MCP Server Helps Across Your Organization
16 real-world use cases for the DataGalaxy MCP Server, organized under two pillars: consuming trusted metadata, and enriching the catalog.
How to Add DataGalaxy MCP Server to Cursor
Step-by-step guide to connecting the DataGalaxy MCP Server to Cursor, including a working JSON configuration example.
DataGalaxy MCP Server Tools
Current reference for every tool the DataGalaxy MCP Server exposes — parameters and behavior, pulled directly from the live server.
How to Add a DataGalaxy MCP Server to Claude Desktop
Connect the DataGalaxy MCP Server to Claude Desktop via the mcp-remote proxy, including a working config and troubleshooting steps.
Quick Start with MCP Server
The 8-step path to your first working DataGalaxy MCP Server connection: enable it, get your token and server URL, and configure your client.
How to connect DataGalaxy MCP Server to Onyx
Step-by-step guide to connecting the DataGalaxy MCP Server to Onyx AI as an Action, with API key authentication.
What is MCP Server and How It Works
What an MCP server is, why organizations use one, and how a request flows from an AI client through DataGalaxy's MCP server and back.
AI functionalities
Automatic description generation with human validation
Generate AI-powered descriptions and summaries for your catalog objects, with human review before they're applied.
Best Practices for Using Description Autogeneration
Practical guidance for Data Stewards and Data Readers to get the most out of AI-powered description and summary autogeneration.
Description Generation by Batch: Automatically Enrich Your Data Catalog
Generate missing descriptions and summaries at scale across a workspace, with admin controls and full history tracking.
MetaBot
Automatic links generation
How MetaBot suggests implementation and usage links by comparing object labels: launching generation, and accepting or rejecting suggestions from the object card or the Suggestions center.
Data classification via MetaBot
Where to find MetaBot's tag and data class suggestions (object card, service panel, suggestions panel) and how to accept, reject and filter them.
MetaBot
What MetaBot is: tag and PII suggestions, how suggestions are generated, the evaluation score and voting system, where to see suggestions, and when to use it.