Protect sensitive data in custom elements

Martin Makarsky

18 minutes

Custom elements enhance your Kontent.ai project by allowing additional functionality. Make them reusable by configuring JSON parameter properties. Bear in mind, though, that all contributors, regardless of their permission levels, can view data in this field, including sensitive information like API keys.

To manage our products, we will use Magento's API

. Because Magento exposes your management API key for your products, we want to provide a content editor with only a read-only product selector. In short, we don't want to allow a content editor to edit, create or delete products.In addition, Magento's API requires an access token while requesting their API which is also sensitive data and should not be shown in your custom element's properties. In this tutorial, you'll learn how to use and protect your sensitive data needed for custom elements by creating a server-side proxy solution to handle your sensitive information.

Server-side proxy

As a proxy, we will configure a simple AWS Lambda function

, which will receive requests from our custom element, add an authentication token, copy headers and query strings, send the request to Magento API, and then resend the response back to our custom element. Since the authentication token is stored within the AWS Lambda function, it's not accessible to contributors on your project.

Communication diagram

In this tutorial, you will:

Create an AWS Lambda serverless function with API Gateway.
Configure the proper host and authentication token.
Send test request to our proxy.

Step 1: Creating an AWS Lambda function

First, you need an Amazon Web Services account. Visit Amazon Web Services

and if you don't have an account, sign up. Once you are logged in, select the data center that best fits your location or purpose in the upper-right corner. Expand All services and click Lambda.

AWS management console — AWS Management Console

On the next screen, click Create a function.

To create the function:

Select Author from scratch
Name your function requestRepeater.
Under Runtime, select Node.js.
Click Create function.

Specifying function name and runtime. — Specify function name and runtime.

On the function's Configuration page:

Click API Gateway trigger from the left-side panel.
Navigate to Configure triggers.
Under API, select Create a new API.
Under Security, select Open.
Click Add in the lower right of the screen.
Click Save in the upper-right corner.

Note: For demonstration purposes, this endpoint will be publicly available and can be invoked by all users.

Be sure to copy the API endpoint to your clipboard because you will need to enter it in Step 3. This URL is the endpoint of our proxy server. Click the requestRepeater Lambda function.

Paste the following code in Function code.

JavaScript

const https = require('https');
const querystring = require('querystring');

/* ========Config Section======== */
const host = process.env.HOST;
const path = process.env.PATH;
const accessControlAllowOriginValue = process.env.ACCESS_CONTROL_ALLOW_ORIGIN;
const accessControlAllowHeadersValue = process.env.ACCESS_CONTROL_ALLOW_HEADERS;

// Bearer token authentization
const bearerToken = process.env.BEARER_TOKEN;

//  Basic authentication credentials   
const username = process.env.USERNAME;
const password = process.env.PASSWORD;
/* ========Config Section======== */

let authorizationHeaderValue;

if (bearerToken || (username && password)) {
    authorizationHeaderValue = bearerToken ?
        `Bearer ${bearerToken}` :
        `Basic ${new Buffer(username + ":" + password).toString('base64')}`;
}

const request = (queryStringParameters, headers) => {

    const requestOptions = {
       host: host,
       path: path,
       port: 443,
       method: 'GET',
    };

    if (queryStringParameters) {
        requestOptions.path = `${requestOptions.path}?${querystring.stringify(queryStringParameters)}`;
    }

    if (authorizationHeaderValue) {
        headers['Authorization'] = authorizationHeaderValue;
    }

    headers['Accept'] = 'application/json';
    headers['accept-encoding'] = 'identity';
    headers['Host'] = host;
    
    requestOptions.headers = headers;

    return new Promise((resolve, reject) => {
        https.request(requestOptions, response => {
            let data = '';
            response.on('data', chunk => {
                data += chunk;
            });
            response.on('end', () => {
                const dataObject = JSON.parse(data);
                response.data = dataObject;
                resolve(response);
            });
        })
            .on('error', error => {
                reject(error);
            })
            .end();
    });
};

exports.handler = (event, context, callback) => {

    const corsHeaders = {
        'Access-Control-Allow-Origin': accessControlAllowOriginValue,
        'Access-Control-Allow-Headers': accessControlAllowHeadersValue
    };

    const repeatResponse = (response) => {
        let multiValueHeaders = {};

        for (const headerName in response.headers) {
            if (Array.isArray(response.headers[headerName])) {
                multiValueHeaders[headerName] = response.headers[headerName];
                delete response.headers[headerName];
            }
        }

        callback(null, {
            statusCode: response.statusCode,
            body: JSON.stringify(response.data),
            headers: { ...response.headers, ...corsHeaders },
            multiValueHeaders: multiValueHeaders,
        });
    };

    const sendError = (error) => {
        callback(null, {
            statusCode: '400',
            body: JSON.stringify(error),
            headers: corsHeaders,
        });
    };

    switch (event.httpMethod) {
        case 'GET':
            request(event.queryStringParameters, event.headers)
                .then((response) => {
                    repeatResponse(response);
                })
                .catch(error => {
                    sendError(error);
                });
            break;
        default:
            sendError(new Error(`Unsupported method "${event.httpMethod}"`));
    }
};

Click Create function and then Save.

Step 2: Configuring your Lambda function

In the Environment variables section, enter the following keys and values:

BEARER_TOKEN – Your Magento access token
HOST – demo1-m2.mage.direct
PATH – /index.php/rest/V1/products

Step 3: Testing the proxy

To test the Magento API via your new proxy (which is accessible through the API Gateway), enter your requestRepeater API endpoint as {Gateway API URL} and then enter the URL in your browser address bar to try the GET request:

HTTP

{Gateway API URL}?searchCriteria[pageSize]=10&searchCriteria[filterGroups][0][filters][0][field]=name&searchCriteria[filterGroups][0][filters][0][conditionType]=like&searchCriteria[filterGroups][0][filters][0][value]=%25watch%25

In our example:

HTTP

https://vpzvj1fspi.execute-api.eu-central-1.amazonaws.com/default/requestRepeater?searchCriteria[pageSize]=10&searchCriteria[filterGroups][0][filters][0][field]=name&searchCriteria[filterGroups][0][filters][0][conditionType]=like&searchCriteria[filterGroups][0][filters][0]

If entered correctly, you will receive a JSON response from the Magento server which was delivered via your new Lambda function.

To learn more about Magento API, see more about our Magento integration

Summarized

When you need to enter sensitive data to your custom element configuration, use an external server-side service like AWS Lambda or Azure Function. After you complete the integration with the service, you get a server-side proxy to handle your sensitive data, request and response headers, and query parameters. You can use this approach with any custom element that works with sensitive information like API keys.