Define the job.
Set a budget and name the places the agent is allowed to spend.
POLICY-FIRST AGENT PAYMENTS
Your agent has a job to do.
Give it a route to pay.
Keep the right to say no.
02 / THE ROUTE
Agents should carry authority that is specific, limited and explainable. Orqrail puts a policy checkpoint between an intention and a payment.
Set a budget and name the places the agent is allowed to spend.
Evaluate the merchant, the amount and the remaining allowance.
Record the decision and its reason. A decline is part of the story, too.
03 / YOUR BOUNDARIES
Change the policy here, then send a request through the terminal. Every control below changes the local demo.
Back to the terminal ↗Only listed merchants can pass the policy.
04 / DECISION TRAIL
| REQUEST | MERCHANT | AMOUNT | DECISION | REASON |
|---|---|---|---|---|
| No requests yet. Run an authorization in the terminal to create your first decision receipt. | ||||
05 / BEFORE YOU BEGIN
No mysterious permissions.
No promises hidden in the small print.
No. This is a working frontend prototype. Budgets, merchants, transactions and decisions are local demonstration data. It does not issue cards or move money.
Add or remove merchants, switch to categories, change a per-request cap, freeze the agent, and run an authorization. The demo evaluates your current rules and writes a decision to the local audit log.
The budget stays unchanged. The audit log records the attempted amount and a reason: frozen agent, merchant restriction, request cap or exhausted budget.
Card issuing, funding, real payment processing, durable audit storage, identity and MCP integration require future backend services. None are connected in this prototype.
No verified token utility or live token is part of this concept. Brand handles and domains are candidates until checked and registered separately.
LET THE AGENT DO THE WORK.
Keep the authority. ↗