For businesses

A free checkpoint before AI agents act on your website.

A defensive layer for the new traffic pattern: AI agents trying to submit, shop, book, and trigger workflows on your site.

The problem

Agent traffic is here. Trust shouldn't be automatic.

01

AI traffic looks like customer traffic

Agents browse, fill forms, and trigger workflows. A bot wall can't answer the real question: should this agent take this action right now?

02

Blocking every agent leaves money on the table

Some agents are legitimate buyers and operators. Distinguish passported agents from anonymous automation instead of treating them all the same.

03

Trust has to be revocable

Every decision is current-state: allowed, needs approval, denied, expired, or revoked. Stop trust before the next risky action runs.

Free checkpoint

Install one verifier before one risky action.

Start with the action where agent traffic can create real cost: checkout, booking, account change, refund, support request, or API write.

01

One protected action

Start with a single route, form, checkout step, booking, or API write.

02

Scoped passport

The agent proves owner, purpose, scope, expiry, and revocation state.

03

Enforceable decision

Your system receives allowed, denied, expired, revoked, or needs approval.

live verifier

POST /v1/verify

agent_idshopping-agent
actioncheckout.submit
scopepurchase:single
expires_in12m
requestcheckout.submit · shopping-agent$42
policypassport.active · scope: purchasefresh
decisionneeds_approval · owner approvedallowed

decision returned

needs_approval → allowed

The site enforces the answer before the action runs. Trust can expire or be revoked later.

Where it goes

Protect the actions that matter.

  • Lead forms and quote requests
  • Checkout, booking, and reservations
  • Account creation and changes
  • Support, refund, and cancellation requests
  • API actions that write data or spend
  • High-volume submissions

First rollout

Start with one risky action, not a platform migration.

01

Choose the action

Pick one form, route, checkout step, or API action where unknown agents create real risk.

02

Add the checkpoint

Place Gardien before that action so the site can ask for a current verifier decision.

03

Decide by policy

Allow trusted low-risk work, require approval for sensitive actions, deny unknown agents.

04

Review the trail

Keep a clean decision record without storing raw customer data everywhere.

Gardien doesn't promise every verified agent behaves perfectly. It gives you a decision point, an audit trail, and a revocation path before the action that matters.

Dashboard

See the checkpoint console a business would use.

The business dashboard is split into simple pages for approvals, protected actions, agents, policy, audit, and settings so teams are not dropped into a crowded control room.

Open business dashboard
Business checkpointLive
Allowed 842
Paused 96
Denied 346

Next step

One website action shouldn't blindly trust AI agents. Start there.