Currently, administrators have no central dashboard to see a live list of which users have active connections to external AI tools (e.g., Claude, GPT) or the ability to selectively disconnect a specific user's connection without revoking the entire domain or app.

User Interest/Use Case: Organization admins need to audit and manage AI tool usage on a granular level to ensure compliance and data security. If a specific user is identified as a risk or no longer requires access, admins need a way to revoke that specific user's MCP authorization without impacting the rest of the organization.