Agent Security › Register
Know which AI agents you have, who owns them and what they're allowed to do. A simple, free workbook for small teams: one place to record your agents and the controls that matter.
Once a business starts using more than one or two AI agents, it becomes surprisingly easy to lose track of what each one does, what it can access and who is responsible for it. Our free AI Agent Register gives you one simple place to record the agents your business is using, and the controls that matter.
A lightweight workbook with three tabs: Your Agents (your working register), Review (automatically highlights entries that may need attention) and Start Here (guidance and examples to help you complete it properly). No macros. No account required. No security expertise needed.
For each agent, record what it does, who owns it, where it runs, how it starts, what it can do, what important systems or data it can access, what authority it has, where human approval is required, whether it runs unattended, how you know its work actually happened, and when it was last reviewed.
The workbook also highlights agents that may need your attention, including agents with no clear owner, undefined authority, missing verification, or unattended operation without evidence of success.
You don't need to catalogue every AI tool someone in the business has ever opened. Start with agents that can take actions, access important information, communicate externally, change systems or run without someone watching them. For each one, ask five questions:
That alone can reveal gaps that are easy to miss when agents are introduced one at a time.
The spreadsheet is deliberately lightweight. It's designed for small businesses and teams with a manageable number of agents that need a clear starting point. If keeping the register up to date starts becoming a job in itself, you've probably outgrown a spreadsheet. And that's useful information too.
The register records what an agent is supposed to be allowed to do. It does not enforce those boundaries. The important limits still need to be enforced in your systems.
If you're unsure what authority an agent should have, use our AI Agent Authority Builder to define it. If you want to identify gaps in the controls around an existing agent, use our AI Agent Security Check.
Start by finding out what you actually have.