Keyri Bouncer:

Fewer Captchas. Happier Users. Less Fraud.

Keyri Bouncer can help your application substantially reduce the surface area for fraud without sacraficing user experience.

Installation

npm i keyri-bouncer --save

Getting Started

Import the library and initialize it:

// Import the library
import { Bouncer } from "keyri-bouncer";

// Instantiate the library to load into shadow-dom
const bouncer = new Bouncer();

// Call the method with your eventType, userId, and your publicKey
let info = await bouncer.interrogate(
    data.eventType, // The type of event: Login, Signup, Visits, Access
    data.userId, // The id of the user in your system
    data.yourPublicEcdhKey // This comes from our dashboard and is used to identify you and as an encryption key
);

// Do something useful with the data
console.log({info});

What Does It Return?

The interrogate method returns an encrypted JavaScript object that contains information about the user's activity and device. Here's a breakdown of the properties you'll find in this object:

• state: Represents the result of the event, depending on your risk tolerance settings. Possible values are "warn", "allow", "deny".

• ipAddress: The IP address of the client.

• userId: The ID of the user in your system, usually their email address. Validate this value to ensure its accuracy.

• deviceId: A unique ID we've assigned to the client's device.

• wagId: A combination of the user's screen size and IP address. Useful when you want to block a specific user on public Wi-Fi, rather than blocking an entire IP address.

• signals: An array that lists suspicious signals detected during the session, e.g., "max_events_per_timeframe".

• trustScore: A score between 0 and 1, based on browser metrics, behavioral analytics, and Bayesian machine learning. A higher score indicates a "good" user.

• changes: An array of changes being made to the user or the device. For example, when a new country is added, or a new IP is registered, these changes will be recorded here.

• event_type: The type of event logged, like "login", "signup", etc.

• deviceAge: The age of the device ID in hours. Since misbehaving device IDs can be blocked, older device IDs are generally more trustworthy.

Here's a typical decrypted response:

{
  "state": "warn",
  "ipAddress": "6.6.6.6",
  "userId": "[email protected]",
  "deviceId": "399c6617-6b9b-4c45-a9fb-6b827cf845c5",
  "wagId": "BxZQPgacdjFm6lKADc3F72Pb5o0=",
  "signals": [
    "max_events_per_timeframe"
  ],
  "trustScore": 0.0750399185765636,
  "changes": [
    {"type": "new_country", "value": "Antarctica"}
  ],
  "event_type": "login",
  "deviceAge": 0.0242025
}