Govern Agent Distribution
Control how agents reach users and how much to trust an agent you did not build — judge third-party trust signals, gate org-built agents through approval, scope who gets which agents, and block the ones that should not be there.
Core capabilities
-
Get a baseline, self-declared security and compliance picture of a third-party agent before adopting it
-
Prefer agents with independent, audited assurance for sensitive or high-governance use
-
Gate org-built agents through approval, scope each agent to only the users who need it, and block the ones that should not be there — all from the admin center
Supporting capabilities 2
-
Keep a single inventory of every agent across the tenant to govern from
-
Approve or block the third-party tools and MCP servers agents depend on