How does the Nivah approval artifact prevent unauthorized purchases?

The approval artifact is Nivah’s spending gate. It keeps research separate from authorization. Nivah can research products, rank options, and explain tradeoffs, but it cannot spend until the shopper reviews an exact approval sheet tied to fresh quote terms.

At the center of that control is a simple rule: no exact approval, no purchase.

What the approval artifact is

In Nivah, the approval artifact is the exact approval sheet that follows a fresh Binding Quote. It binds the decision to specific terms, not to a vague intent.

That sheet covers the purchase details that matter most:

product
seller
variant
quantity
destination reference
shipping method
tax
delivered total
recurrence terms
expiry
route, when applicable

This matters because a different seller, price, package size, or delivery setup is not the same purchase. Nivah treats those changes as new terms, not as silent permission.

How it blocks unauthorized spending

The approval artifact prevents unauthorized purchases by making spending a deliberate, checked action instead of a side effect of conversation or inference.

The control flow is strict:

Nivah first surfaces a capability-checked offer.
Market References can be useful for research, but they are non-selectable if Nivah cannot prove a safe checkout route.
Nivah fetches a fresh Binding Quote.
The quote reflects the current product, seller, delivery, and total.
Nivah shows the exact approval sheet.
The shopper sees the exact terms before anything can move forward.
The shopper must confirm the approval control.
Voice alone cannot spend. A spoken “yes” is not enough.
Only then does Nivah open the separate Prava authorization experience.
Prava is the secure authorization layer.
After authorization, Nivah may execute only a route that has already passed deterministic capability checks.

That sequence stops accidental, inferred, or partial permission from becoming a purchase.

Why exact terms matter

The approval artifact is strong because it is specific.

If the seller changes, the prior approval no longer stands.
If the price changes, the prior approval no longer stands.
If the package size changes, the prior approval no longer stands.
If the shipping or destination reference changes, Nivah must refresh the quote and ask again.

This is why Nivah uses phrases like fresh Binding Quote and exact approval. The approval is tied to one exact transaction, not to a general desire to buy something similar.

A practical example: if a shopper is renewing razor cartridges and the merchant changes the pack size or unit price, Nivah stops. The shopper can approve the new terms, skip once, change the cap, switch products, or pause the intent. Nivah does not silently continue.

How Nivah keeps authorization separate from payment

Nivah does not collect or store card numbers, CVVs, OTPs, passkeys, or one-time payment credentials. That is another layer of protection against unauthorized purchases.

It also means that:

a Prava session is not a purchase
a payment session is not a purchase
a credential-ready status is not a purchase

A purchase is only recognized after merchant evidence is reconciled. If the merchant result is unclear, Nivah reports Needs verification. It does not guess.

That separation keeps approval, authorization, and merchant success from collapsing into one step.

What the approval artifact does not do

The approval artifact does not grant blanket spending power. It does not turn inference into permission. And it does not make every discovered merchant executable.

Nivah distinguishes these states clearly:

Autonomous Offer: capability-checked and eligible for a fresh quote and approval
Market Reference: useful for research, but non-selectable
Controlled sandbox/test route: available for public demonstrations, but not a live merchant purchase
Needs verification: outcome is unknown and must not be treated as success

That distinction is important. It keeps public market evidence separate from routes Nivah can safely execute.

The practical result for shoppers

For shoppers, the approval artifact creates a predictable boundary.

You can let Nivah do the research.
You can let it compare options against your constraints.
You can let it prepare the exact terms.
But you still decide whether the purchase goes forward.

That design reduces unauthorized spending in three ways:

it blocks spending before exact approval
it requires a second, separate Prava authorization step
it refuses to treat unclear outcomes as success

The result is useful automation without giving up control over spending.