Overview
TrueFoundry’s AI Gateway now offers Collibra as an official TrueFoundry’s remote MCP server, meaning any Collibra instance can be connected to TrueFoundry’s MCP Gateway through OAuth2 with no custom server deployment required. For joint customers, this turns Collibra’s platform into something AI agents can actually query and reason over, instead of information locked behind a UI that only humans browse. Because the Gateway runs the OAuth2 Authorization Code flow with PKCE, each user authorizes their own Collibra account and every tool call still runs under that person’s existing Collibra permissions, so governance and access boundaries carry through automatically rather than requiring a new permissions model for AI access.
The practical benefit is that organizations already standardized on TrueFoundry for managing LLM and agent infrastructure get a single, centralized, and observable path into their governed data assets. Rather than every team wiring up its own connection to Collibra, agents authenticate once through TrueFoundry’s unified endpoint, and admins get one place to manage collaborators, enable or disable individual tools, apply guardrails, and monitor usage (invocation counts, latency, error rates) across every MCP server in the environment, Collibra included. For Collibra customers specifically, this means their investment in data governance and cataloging can be extended into agentic workflows with full auditability: every tool call is traced with caller identity, inputs, and latency, so data stewards retain visibility into which agents touched what platform data and when. In short, the integration lets customers combine Collibra’s trusted data context with TrueFoundry’s agent orchestration and control plane, so AI applications can ground their answers in governed, well-described data without sacrificing the access controls and oversight that made that data trustworthy in the first place.
