# Agent permission card

Fictional example: the invoice-matching agent of the finance team at Lumen, an invented software company.

One card per shared agent: what it’s for, what it can reach and do, what needs approval, its limits and how to stop it.

## Agent and purpose

Name the agent and the one job it does.

Invoice matcher: matches supplier invoices to purchase orders.

## Owner

Name the person responsible for the agent and its card.

The accounts payable lead.

## Can read

List the systems and data it can read.

The supplier invoices inbox (read-only), purchase orders and supplier master data.

## Can do on its own

List the actions it takes without approval.

Match invoices to orders and flag mismatches.

## Needs approval

List the actions that wait for a person.

Marking an invoice as approved for payment.

## Never

List the actions it must never take, enforced in its permissions.

Change supplier bank details, make payments or email suppliers.

## Limits

Set limits on money, recipients, volume and time.

Invoices up to €10,000. It stops after 200 invoices a day.

## Reads from outside

List the untrusted content it reads, such as emails or web pages, and how that’s handled.

Supplier emails and PDFs. Instructions inside them are ignored, and anything asking for a change is flagged for a person.

## How to stop it

Say who can pause it and how.

Any finance team lead, with the pause switch in the automation settings.

## How to undo its work

Say how to reverse what it did.

Matches can be reset from the audit log.

## Review date

Say when the card and its permissions are checked.

Quarterly, and whenever a permission changes.

## Review

Owner: 
Agreed by: 
Date agreed: 
Next review: 

---

From The RUAI Standard (2026 edition), developed by Founderz in collaboration with Microsoft. https://responsibleai.founderz.com/toolkit/team-tools/agent-permission-card
Licensed under CC BY 4.0: https://creativecommons.org/licenses/by/4.0/