How to secure AI agents that use local MCP servers

​Teams are already running agents locally on laptops, against real infrastructure, often over-permissioned and without any oversight from security or IT. That's a risk to the business, not just to the local machine.

Join us for a practical session on how to give developers the freedom to run local MCP servers, while giving your IT and security team the access control and visibility they need.

​Tim Stacey is the Co-Founder & CTO of Barndoor, the control plane for agentic AI, where he builds the access control and governance infrastructure that lets enterprises run AI agents and MCP servers securely.

​You'll learn:

• ​Why local agents, running on laptops and often over-permissioned, are a growing blind spot for security teams

• ​How to let developers run MCP servers locally without slowing them down or getting in their way

• ​How to apply access control policies to local MCP usage, enforced behind the scenes

• ​How to get visibility into who's using which local MCP servers, and what they're doing

• ​How it works in practice: a real agent, enforced live with Barndoor CLI

​Who is this for?

• ​CISOs & Security Leaders: Close the blind spot on local agent access before it becomes an incident

• ​Engineering Managers: Let your team move fast on local MCP without losing oversight

• ​Platform & SRE Teams: See a working pattern for enforcing access control on local agents

• ​Developers: Understand what guardrails look like without them getting in your way

​Event Details

• ​Date: 8/11/2026

• ​Time: 11am PT / 2pm ET

• ​Location: Live on Livestorm (Yes, we'll send a recording)

​Save your spot today.