Skip to content

Commerce / Security

Access, controls and grounded protection.

Flur Commerce is built around deliberate permissions, scoped roles, and checkpoints that require a real decision before consequential actions proceed.

Access control

Scoped access for your team

Team roles limit what each person can see and do. Assign each team member the scope appropriate to their responsibilities. Avoid sharing a single login across multiple people; individual accounts allow you to track who took which action.

Owner

Full access to all settings, users, billing, and commerce data.

Manager

Manage products, orders, and inbox. Cannot modify billing or user roles.

Cashier

Process payments and view orders. No access to settings or catalogue editing.

Approval workflows

Controlled consequential actions

Approval workflows require a deliberate decision before a payment-related action progresses. No offer advances to payment automatically unless the buyer has set up explicit Automatic Approval rules with defined bounds — a price ceiling, an approved seller list, or a time constraint. Outside those bounds, manual confirmation is required.

Manual Approval

Every offer is reviewed and explicitly approved by the buyer before it can progress. No automatic action occurs.

Automatic Approval (bounded)

The buyer defines conditions. Only offers that exactly match those conditions progress automatically. Anything outside the defined bounds requires manual review.

AI flows

Koreshield in Flur AI flows

Koreshield is a checkpoint in relevant AI-assisted commerce actions. When Flur AI participates in a buying or selling flow, Koreshield evaluates the action before it executes. Flur retains control of permissions and commerce actions; the AI does not act outside the scope it has been given.

What Koreshield does

  • +Acts as a checkpoint before AI-assisted commerce actions execute
  • +Ensures AI actions remain within the permissions and scope assigned by the user or seller
  • +Flags actions that fall outside defined bounds for manual review
Koreshield is a control mechanism, not a total protection guarantee against all incorrect AI outputs or fraud. It reduces the scope of unintended AI actions; it does not eliminate all risk.