The CDGC Metadata Search server allows your AI agents to perform natural language searches to discover data assets, understand their context, and retrieve essential metadata from Cloud Data Governance and Catalog (CDGC).
Use this MCP server to ground your LLM responses, ensuring that your AI only works with governed, approved, and correctly-defined enterprise data assets, preventing hallucination and data misuse.
Key capabilities
This MCP server offers the following key capabilities:
•Metadata retrieval. Returns crucial metadata for grounding responses, including the asset name, type, description, and unique ID.
•AI-powered classification. Provides an understanding of data across the organization using CDGC's automated classification, for example, as personally identifiable information (PII) or financial data, as well as its associate business terms.
•Contextual grounding. Provides your AI models with the necessary enterprise context to generate trustworthy and relevant insights.
Use cases
Use this MCP server to address the following use cases:
•Data governance and compliance. Quickly identify and report on tables containing sensitive data like PII.
•Data discovery. Discover the source or related data assets to understand a data element's full context.
•LLM response grounding. Provide your AI with the exact name and description of a data asset before it executes a data query.
•Documentation and onboarding. Quickly find and summarize technical and business descriptions of key assets for new users.
Tools
The following table describes the tools available on this MCP server:
Tool
Description
get_asset_details
Retrieves the complete, specific metadata details for a data asset. This includes lineage, profile, and full descriptions.
search_metadata
Allows AI agents to query the Informatica data catalog for enterprise data assets and returns a list of matching asset identifiers and core attributes.
Note:
In AI Agent Engineering, the MCP server tools are called "actions."