Loading...

Vetting API

Overview

The E-HAWK Vetting service provides real-time reputation and risk analysis. The Vetting API is used to call the vetting service during sign-up, account updates, logins, or other user interactions. Please follow this guide for implementation instructions.

Endpoint (6.4)

https://api.ehawk.net/

Using HTTPS RESTful calls, your online forms or back-end systems query our service about risks levels associated with the information entered by users.

The API requires HTTPS POST and Content-Type: application/x-www-form-urlencoded

For all API calls:

  • Data is sent with name/value pairs keyword=value
  • Properly encode all values within the HTTPS POST
  • All values must be in UTF-8 character sets
  • x-www-form-urlencoded does not support boolean, so all true/false settings must be strings
As an example, calling the API with CURL (see PHP example and POSTMAN testing below):

curl -X POST -H Content-Type:application/x-www-form-urlencoded -d 'apikey=your_apikey&ip=10.1.1.1&email=me@test.com&country=us' https://api.ehawk.net/

Keywords

The Vetting API will analyze the following input values to calculate a Risk Score. Every API call requires the apikey and either an ip, email, domain, or link value. However, the more keyword and value pairs provided, the better and more detailed the analysis. API call Keywords must be lowercase. Please follow value formats listed below.

Keyword Value and Format
apikey Your Vetting API KEY (required)
ip IP address of the user connecting browser. IPv6 supported but IPv4 recommended.
email email address (name@example.com)
talon See EHawkTalon.js for device fingerprinting implementation
phone US and Canada: 10 digit format XXXXXXXXXX
International: "+" AND country code AND number, ex: +33143542331 (France phone)
street Street, PO box, location
city City, town, or village
state State, province, or area. US and Canada must be two-letter lowercase code (example tx=Texas). Other countries just use the actual state if available.
postalcode Postal code or zip code
country Two letter lowercase country code (example us=USA). Be sure to send country value if sending any address data.
domain User’s verified domain.
Not necessarily the domain of their email, but the domain they own or will be sending email from. Domain should be linked to their business or “From” address of their emails. Example “ehawk.net”. Do not use URL (no http://, just “domain.com”) If no domain then send the email domain as domain, so jim@aol.com would send domain=aol.com
website URL of the users website (http://www.mysite.com). Send either keyword 'domain' or 'website', but not both. If both sent, we just process domain.
link URI/URL of embeded link to check for possible spam, phishing and other risks. Use fully qualified URLs such as https://www.example.com/webpage.html for links embedded in messages asking users to click.
firstname First name (No prefix such as 'Dr.'. No middle initial.) Do not send company names.
lastname Last name (No suffix such as 'Jr.' or 'III') Do not send company names
username The unique username or userID of the account on your systems or platform. Truncated at 50 characters. Do not use email for username.
useragent The User-Agent string provided by the browser
referrer the referring URL provided by the browser during form completion
lead_id ID of lead in your lead management system. Max length: 50 characters.
lead_source ID, name, or URL that is providing the leads. Max length: 250 characters.
sub_id Lead Source Sub ID. Max length: 250 characters.
campaign_id Campaign ID or name. Max length: 250 characters.
bank_name Bank name
account_number Bank account number
routing_number Bank routing number. In US this is the ABA.
driving_
license_number
Driver's license number. US only.
driving_
license_state
Two-letter lowercase code for driver's license issue state. US only.
birth_date Birth date in format YYYY-MM-DD. Required for social_
security_number testing.
social_
security_number
Social security number in format 123456789. USA only.
revet true (string not boolean)

IMPORTANT! The vetting service by default uses activity testing to tag similar data from multiple sign-ups, bots, etc. If you want to manually vet a user again within a short period of time, you should add revet=true to the API call, so it bypasses the activity tracking engine. Otherwise our system will think the person is trying to sign-up multiple times and start marking that user as high risk. The service will still display any previous activity results but will not increment counters or change activity scoring. Default is false. Must be sent as string, not a boolean.
timeout false (string not boolean)

If timeout=false then the vetting engine will perform important, additional testing on domain, email, phone and other areas. If you can wait a few seconds for the API response, then add this to your API calls, as the extra data is valuable. Must be sent as string, not a boolean. Admins can set default as 'false' in the portal so all API calls use 'false'. Default is 'true'.

Talon (6.0.8)

The API is supported with a JavaScript device analysis technology. Incorporating this into your forms and sign-up process increases accuracy and provides extra data for E-HAWK to analysis.

When the script runs it will update the value in the hidden field "talon6" in your form. In the form POST you will get a name value pair talon = value. This value (JSON payload) you send to the API together with the other data such as IP, email, etc.

Example implementation where you can see the Talon Generation:

  
<!DOCTYPE html>
<html>
<head>
    <meta charset="utf-8" />
    <title>EHawk Talon</title>
</head>
<body>
    <input type="text" name="talon6" id="talon6" value='{"version": 6, "status": -1}'>
    <script type="text/javascript">
        var eHawkTalonSettings = {
            cookie: {
                SameSite: 'Lax',
                Secure: false,
            },
            bind: {
                OutId: 'talon6',
            },
        };
    </script>
    <script src="path_to_your_talon_js_6_file.js"></script>
</body>
</html>

The bind=>OutID value should match the form element id name. When using on your webpages, change the input type to "hidden" and the talon payload will be part of the form POST but not viewable on the form.

Download Talon JS in the portal (Account=>Settings, bottom of page) and add to your JS files or CDN.

Additional Settings

Talon 6.0.8 and higher supports options for autoLoad, cookie key name, and async calls for advanced installations.

  • autoLoad: true or false. When set to false, make sure to call Talon.eHawkTalon();
  • when wishing to execute code after Talon is completed (previously oncomplete callback in Talon v5), use await when calling Talon.eHawkTalon(), to ensure talon has finished loading
  • to customize the name of the cookie key use cookieName

<!DOCTYPE html>
<html>
<head>
     <meta charset="utf-8" />
     <title>EHawk Talon</title>
</head>
<body>
    <input type="text" name="talon6" id="talon6" value='{"version": 6, "status": -1}'>
    <script type="text/javascript">
        var eHawkTalonSettings = {
            cookie: {
                SameSite: 'Lax',
                Secure: false,
            },
            bind: {
                OutId: 'talon6',
            },
            autoLoad: false,
            cookieName: 'talon6',
        };  
    </script>
    <script src="path_to_your_talon_js_6_file.js"></script>
    <script type="text/javascript">
        // When autoLoad == false, call Talon manually,
        // EITHER by calling it directly:
        Talon.eHawkTalon(); 
        
        // OR by calling it async, but not both:
        async function execTalon() {
            await Talon.eHawkTalon();
            const tal = document.getElementById('talon6');
            console.log('talon val', tal.value);
            // execute custom code after Talon has finished.
        } 
        execTalon();
    </script>
</body>
</html>

Please contact support@ehawk.net for any additional info on the Talon.

Response

The output returned by an API call is in JSON format. There are two response formats that can be selected in the Portal Settings area: Format 2 (new and compact) and Format 1 (original).

{
  "version": "6.4",
  "transaction_id": "5808be52e52542.29287853",
  "status": 0,
  "error_message": "",
  "score": {
    "risk": -100,
    "type": "Very High Risk",
    "total": -498
  },
  "area": {
    "ip": -80,
    "email": -155,
    "phone": 5,
    "domain": 0,
    "geolocation": -24,
    "community": -139,
    "combo": -80,
    "fingerprint": -25
  },
  "fingerprint": {
    "id": "15da7f7b71f17e237ff9197e19802399",
    "hits": 4
  },
  "risk_hits": {
    "ip": [
      "Proxy - Anonymous",
      "Spam Blacklist"
    ],
    "email": [
      "Disposable",
      "Emaildomain Age 1 to 5 Days"
    ],
    "geolocation": [
      "IP vs location > 6,000 miles"
    ],
    "activity": [
      "4 Repeats"
    ],
    "combo": [
      "Name no Vowel and Email Disposable"
    ],
    "community": [
      "Fraud Email",
      "Cyber Crime IP"
    ],
    "combo_details": [
      {
         "name":"Name no Vowel and Email Disposable",
         "score":-80,
         "details":"filter:Name:No Vowels AND filter:Email:Disposable"
      }
      ]
  }
}

A Sample JSON format and areas:

Name Type Description
version string API version
transaction_id string API transaction ID
status number status code. See codes below
error_message string blank if no error, or message of error
score array risk number between -100 and +100
type string of the Risk Level
total number of the total score of all areas
area array each area tested with the score as a number
fingerprint array id string value of fingerprint
hits number of times seen in API calls
risk_hits array area arrays with all the hits for each area

There can be additional data in the JSON depending on service levels and features. Please make sure to properly parse the JSON response to allow for additional data. The key structure above will not change.

{
  "version": "6.4", 
  "transaction_id": "5808be52e52542.29287853", 
  "status": 0,  
  "error_message": "",  
  "score": [     
    "Risk Score",
    -100,       
    "Very High Risk"  
  ],
  "scores": {   
    "ip": [     
      "total",
      -250
    ],
    "email": [    
      "total",
      -135
    ],
    "community": [  
      "total",
      -80
    ],
    "combo": [
      "total",
      -80
    ],
    "geolocation": [ 
      "total",
      -65
    ],
    "activity": [  
      "total",
      -65
    ]
  },
  "details": {
    "score_total": -675,
    "fingerprint": "12e30087109e44e4fd4",
    "fingerprint_hits": 3,
    "ip": {
      "score_details": [
        "Proxy - Anonymous",
        "Bots, Drone, Worm, Proxy, TOR",
        "Spam Blacklist"
      ]
    },
    "email": {
      "score_details": [
        "Disposable"
      ]
    },
    "geolocation": {
      "score_details": [
        "IP vs location > 6,000 miles"
      ]
    },
    "activity": {
      "score_details": [
        "4 Repeats"
      ]
    },
    "combo": {
      "score_details": [
        "Name no Vowel and Email Disposable",
        "IP no PTR and Email Many Periods"
      ],
      "combo_details": [
        {
          "name": "Name no Vowel and Email Disposable",
          "score": -40,
          "details": "filter:Name:No Vowels AND filter:Email:Disposable"
        },
        {
          "name": "IP no PTR and Email Many Periods",
          "score": -40,
          "details": "filter:IP:No Reverse PTR AND filter:Email:Contains Many Periods"
        }
      ]
    },
    "community": {
      "score_details": [
        "Fraud Email",
        "Cyber Crime IP"
      ]
    }
  }
}

The JSON format and areas are:

Name Type Description
version string API version - changes with each release
transaction_id string API transaction ID
status number status code
error_message string blank if no error. Message of error.
score array three elements. Item 1 is a label. Item 2 is the Risk Score (number between -100 and +100), item 3 is the Risk Type string
scores array arrays of each scoring area. Array key is area. Item 1 is a label of 'total'. Item 2 is a number value of the area score.
details array score_total number sum of all scoring
fingerprint string value
fingerprint_hits number of time fingerprint seen by system
area hits if risk hits, then they are listed for each area under score_details array

There can be additional data in the JSON depending on service levels and features. Please make sure to properly parse the JSON response to allow for additional data. The key structure above will not change.

Status Codes

The status will be a value (0 = OK) or an error code and a corresponding errors array. Your script should check for status = 0 for a valid API call and response. As an example, an API call with an invalid apikey would return

{
 "version":"6.4",
  "transaction_id":"123456789",
  "status":-3,
  "error_message":"Invalid API key"
}
Status Description
0 Success - no errors
5 API requires either IP, email, or domain inputs
or 'IP input error' (if IP required for API key)
-1 network congestion (reduced fidelity)
-3 Invalid API key
-4 Account Suspended
-6 IP not in ACL
-8 illegal content (data not sent as UTF-8 character sets)
-12 Out of Vetting Credits
-13 Exceeded Max API Call Rate
-99 Database connection error (System issues)

Sample Code

First, create a web form (form.php) that captures information to send to the API. Here we capture some data such as first and last name and email. We also capture in hidden fields IP, useragent, and talon. Note that the POST values need to match the form name values. When submitted, the form action sends the names and values to 'process.php' that calls the API and processes the scoring and response.

Typically our API responds in about one second. If our services need to contact SMPT or Web servers that are on slow connections, the API can take longer. In the sample code below we set the CURL timeout to 10 seconds. We also recommend adding a queue in your code to call the API with the timeout data again after a few minutes and adding revet=true.

Postman Example

You can test the Vetting API using Postman. Make sure to set post variables as Body and x-www-form-urlencoded.

Top