Use Trust3 IQ MCP Server with AI Agents¶
The MCP (Model Control Protocol) Server provides a standardized interface for AI agents to interact with data infrastructure and models. This guide explains how to configure any AI agent or client to communicate with an Trust3 IQ MCP server.
MCP Client Configuration¶
Your AI agent or client will need to configure the MCP connection settings. Below is a generic configuration structure that provides the configuration needed for the MCP-compliant client:
| JSON | |
|---|---|
The above snippet needs to be part of a bigger JSON file which houses all your MCP servers. A fully configured MCP Servers config would look like this:
| JSON | |
|---|---|
Configuration Fields¶
trust3-iq-mcp-client: Root configuration object that can be customized with your preferred nametransport: Communication protocol (use "streamable_http")url: MCP server endpoint URL with your host address and port. Default port is8000headers: HTTP headers for authentication and response formatAccept: Response formats the client can handle (supports both JSON and Server-Sent Events)x-space: Name of the IQ space to use (must be created in Trust3 IQ UI first)x-api-key: Authentication key for accessing the MCP server. Please contact with Trust3 team for this.timeout: Maximum time in seconds to wait for server response, Default is600seconds
Important: The
x-spacevalue must correspond to an IQ space that has been created in the Trust3 IQ UI. You need to first create your IQ space through the Trust3 IQ interface, then use that exact space name in your configuration.