What's New > October 2026 > New features and enhancements
  

New features and enhancements

The October 2026 release of Informatica MCP servers and AI Agent Engineering includes the following new features and enhancements.

New MCP servers

The following shared MCP servers are added in October 2026:
Catalog Asset History
Lets you investigate how enterprise data assets change over time in Data Governance and Catalog.
Data Access Management
Lets AI agents create, review, and publish data access policies, which filter data, mask data, and control who can access data while following your organization's data protection rules.
IDMC Asset Documentation
Gives AI agents read-only access to asset metadata, so that they can generate technical documentation.
For more information about shared MCP servers, see Shared MCP servers.

Enhanced MCP servers

In October 2026, we’ve enhanced the following shared MCP servers:
Catalog Discovery server
We added the get_asset_lineage tool to enable AI agents to trace asset lineage in Data Governance and Catalog
We also added the Describe Asset type resource that returns the producer schema of a governance or technical asset type.
Catalog Enrichment server
We added the following tools that enable AI agents to enrich and curate enterprise data assets in Data Governance and Catalog:
The existing update_asset_business_metadata tool lets you update custom attributes for technical assets.
Connection Management server
We added the following tools that enable AI agents to browse objects, discover metadata, and preview data in Data Integration connections:
For more information about shared MCP servers, see Shared MCP servers.

MCP Monitoring

The Monitor page in AI Agent Engineering allows you to track the health, performance, and usage of MCP servers that are running or have run in your organization.
For more information about monitoring MCP servers, see Monitoring MCP servers.

MCP server registry

Informatica hosts an MCP server registry that provides a searchable catalog of its shared MCP servers at the following endpoint:
https://hdm.<cloud provider>-<region>.informaticacloud.com/mcp-registry
Use the MCP registry to discover Informatica shared MCP servers without logging in to IDMC and opening AI Agent Engineering.
For more information about the MCP server registry, see MCP server registry.

OAuth client scopes

When you create an OAuth client that uses the Authorization Code flow, you configure the OAuth client scopes. OAuth client scopes define the high-level permissions that an OAuth client is allowed to exercise on an AI agent’s or user’s behalf when it interacts with an IDMC MCP server. For example, a client that has only the Read (mcp:tools:read) scope can't call tools that modify data or objects and can't call tools that execute jobs or processes.
Authorization Code (OAuth 2.1) clients created before the October 2026 release inherit the following OAuth client scopes:
For more information about OAuth client scopes, see OAuth client scopes.

MCP server resources

Shared MCP servers can now contain resources that provide read-only, contextual information to the AI like text or markdown files, zip files, or API responses. You can view information about an MCP server's resources on the Resources tab of the MCP server details page.
For more information about MCP resources, see MCP server details.

Lifecycle states for MCP tools

The MCP Servers page and the MCP server details page for a shared MCP server now list the number of tools with in the preview, active, and deprecated states. When you view details for a tool, the tool's lifecycle state is displayed.
For more information about MCP tool lifecycles, see MCP server details.

MCP server connection information

The Configuration tab on the MCP server details page for each shared MCP server displays connection information for each authorization flow: generic, Client Credentials, and Authorization Code.
For more information about connecting to shared MCP servers, see Connecting to an Informatica MCP server.