CMS for NodeJS: Using Crafter CMS Javascript SDK on the Server with Node JS

In this blog, we will focus on the language-specific bindings for Javascript, the Crafter CMS Javascript SDK applied on the server side with Node JS. A deep dive on the Javascript SDK and architecture can be found in this blog post: Introducing Crafter CMS Javascript SDK

Pre-requisites for Using the Library

  • NPM / Yarn installed
    • The SDK releases are published to Node Package Manager (NPM.)  Example:
    • NPM will help you easily download the proper dependencies.  You can all so use NPM or Yarn to build and execute your application.
    • In our example, I will use Yarn to manage my package.json file and to assist me in building my application.
  • Access to a running Crafter Engine instance and site.

Building a Simple Javascript Application

The point of our application in this blog is to demonstrate how to include and use Crafter CMS’s Javascript SDK in an application and to use it to retrieve content from a running Crafter Engine instance.   We’re going to keep the application extremely bare bones in order to illustrate that this library is independent of any specific frameworks or platforms such as Node.js, React, Angular, Vue etc.

For those looking to get right into the meat of the integration, use of the SDK starts at Step 4.

Step 1: Create a basic structure and build for the app

The first thing we want to do is to create a directory for our application source.  I will call my directory simple-node. Inside that directory add a file called package.json with the following content:

 "name": "simple-node",
 "version": "1.0.0",
 "description": "Simple Node JS Server that will use Crafter CMS Javascript SDK",
 "main": "server.js",
 "scripts": {
 "test": "echo \"Error: no test specified\" && exit 1"
 "author": "",
 "license": "ISC",
 "dependencies": {
 "@craftercms/classes": "^1.0.0",
 "@craftercms/content": "^1.0.0",
 "@craftercms/models": "^1.0.0",
 "@craftercms/search": "^1.0.0",
 "@craftercms/utils": "^1.0.0",
 "express": "^4.16.3",
 "url-search-params-polyfill": "^5.0.0",
 "xmlhttprequest": "^1.8.0"
 "proxy": {
 "/api": {
 "target": "http://localhost:8080"
 "/static-assets": {
 "target": "http://localhost:8080"

This file describes our application, how to build it and what is needed to build it.  In the content above we can see:

  • The application metadata.
  • The application dependencies we are requiring are the various Crafter CMS Javascript SDK dependencies ( “@craftercms/content”: “^1.0.0”, “@craftercms/search”: “^1.0.0”, etc.)
  • The application build/dev dependencies that help us compile the application for distribution (xmlhttprequest”: “^1.8.0.)
  • Configuration for your Crafter CMS endpoint.

Step 2: Perform an initial build

Run yarn in your application directory.  This will download all of your dependencies.

Step 3: Create an application Javascript

Note that in our package.json we’ve listed server.js listed as the main for the application.

Create a server.js file and place the following content in the file:

console.log('server is starting');

var express = require('express');
var app = express();
var server = app.listen(3000, listening);

app.get('/', test);

function listening() {

function test(request, response) {
 response.send('Hello World!');



Now run the application by typing: node server.js

Finally, access your application via the browser at http://localhost:3000


Step 4: Calling Crafter CMS API’s with our application

Now that we have a very basic, working application it’s time to use the Crafter CMS Javascript SDK.

Crafter Engine Instance

In order to use the Crafter CMS APIs, you will need a running Crafter Engine instance with a project set up that you have access to.  It’s easy to get started. Here are some helpful links on getting things set up:

Import the APIs into your application javascript

Add the following code at the very top of your server.js script file.  Here we are declaring the classes we will use and indicating where they are imported from (the SDK.)

var classes = require('@craftercms/classes'),
    content = require('@craftercms/content'),
    crafterConf = classes.crafterConf,
    ContentStoreService = content.ContentStoreService;

XMLHttpRequest = require("xmlhttprequest").XMLHttpRequest;

Configure the SDK for your Crafter Engine endpoint base URL (server name, port etc)

Now that you have declared the Crafter CMS SDK services you want to use you are almost ready to start retrieving content.  Before we can, we need to make sure the SDK APIs know where Crafter Engine “endpoint” / server is and what site to pull the content from.  To do that we have two options:

  • We can create a configuration object and pass it with each call
  • Or we can explicitly supply the endpoint configuration with each call

Since we’re going to use the same endpoint and site, we’ll set up a configuration object and pass it when we call the APIs.  To configure the endpoint we add the following code to our script after the import statements:

 baseUrl: 'http://localhost:8080',
 site: 'editorial'

The configuration above states that our API calls will reference the site in Crafter Engine that has the “editorial” ID. By default, the configuration assumes that Crafter Engine is running locally on port 8080. To change this, add the following:


 site: 'editorial',
 baseUrl: "https://mydotcomsite"

If you want the system to assume that the app and the API endpoints are on the same server you can set base URL to an empty string (“”)

Call a Crafter CMS API and get content

Make the code changes:

Now that the library is imported and configured, we’re ready to use it.  Let’s add the code to our example to make a request for content and put it on the screen for the user.   Below is the entire server.js script for our application with changes:

console.log('server is starting');

var express = require('express');
var classes = require('@craftercms/classes'),
 content = require('@craftercms/content'),
 crafterConf = classes.crafterConf,
 ContentStoreService = content.ContentStoreService;

XMLHttpRequest = require("xmlhttprequest").XMLHttpRequest;

var app = express();
var server = app.listen(3000, listening);

app.get('/', test);

 baseUrl: 'http://localhost:8080',
 site: 'editorial'

function listening() {

function test(request, response) {
 .subscribe(item => {
  var htmlTitle =['hero_title'];
  response.send('Hello World: ' + htmlTitle);


What we’ve done is replace the code in the test function that was hardcoded to write “Hello World” on the screen for the user with the code that calls Crafter CMS for content and write the retrieved content on the screen instead.

Let’s take a closer look at the API call; ContentStoreService.getItem

  • The first parameter is the path of the content item we want to return.  The example uses the home page of the editorial site.

Test the application

After you have made your changes, stop the Node server (Ctrl+C) and restart it. Once it’s restarted, reload your application in the browser to see content coming from your Crafter CMS site:


Congrats!  Your application’s content is now externalized and managed in the CMS.

Step 5: Update your content using the CMS

Now that our code is complete our content authors can update the content at any time via the CMS and we don’t need to rebuild or redeploy the application.  Let’s use Crafter Studio to update the content for our app.   You can find detailed how-to help here:  Creating and working with your First Crafter Site

In Crafter Studio, find the content you want to edit

In our case, the content we’re consuming is the hero title.  Let’s make an edit to that content:
In your site’s content preview, turn on the editing pencils and click the pencil for the Hero Title (as shown above.)

Note that the author has a preview of the content.  This is something very different and ultimately much better than what you see with other Headless CMS systems.

With Crafter, it’s always possible to provide the user with a preview of the content they are editing regardless of the front end framework.  Preview capability is key because it makes editing content easier for authors.

Make a content change and save

After you click the pencil for the Hero text make the change and then click the Save and Close button at the bottom of the form.  You will see your update in the CMS right away.

Refresh the example application

Wow! Those changes in the CMS look great (if I do say so myself :D!)  Now it’s time to see the changes in our application.  Simple, just refresh your application in the browser. No code updates required!

It’s really just that simple!  A few lines of code allows us to create applications with content, configuration and other strings that we can update at any time without needing to update or redeploy our application code.

What about the full API, workflow and publishing?

This blog is meant only to cover the basics of using the Crafter CMS SDK with Node JS. A deep dive on the Javascript SDK and architecture can be found in this blog post: Introducing Crafter CMS Javascript SDK.


Crafter CMS’s Javascript SDK gives front end and full stack developers who work in Javascript and Javascript related technology a native programming API to work with that makes building content-rich applications much easier to do.

When we stop and think about it, almost every application has content in it.  If that content is easy to update and updating the content can be done by non-technical users at any time without a build or deployment everybody wins:

  • The content in the application is richer and much fresher
  • Developers spend time developing rather than making non-code related updates
  • Removing non-feature related deployment distractions gives us more time and makes our teams more agile.
  • Our business users can make their changes at any time which makes the business more agile.

Headless CMS capabilities provide non-technical users with the ability to update, workflow and publish an application’s content at any time without the involvement of developers and IT.

Crafter CMS’s headless capabilities go beyond traditional headless CMS systems by providing authors with an in-context preview for content editing and giving developers the freedom to use whatever front-end technology they want.

CMS for SPAs (2 of 4): Save the Content Authors

In Part 1 of the series CMS for SPAs: Are Single Page Applications and Headless CMS a Slam Dunk?  we looked at the trend toward Single Page Applications (SPA) and Headless Content Management Systems (CMS) in general. SPA applications are becoming the de facto way to build web-based applications and sites. Headless CMS is a decoupled, API-first approach to content management that aligns neatly with SPA architecture. Further, the decoupled nature of the architecture makes content more reusable (multi-channel) and completely divorces development from the CMS allowing for greater freedom and agility for developers. While these can be seen as major wins, several developer and DevOps issues remain while others have been created. Further, most of the headless CMS solutions available today set content authors back nearly 20 years in terms of content editing and workflow tooling and practices.

In this installment, we’ll focus on the various ways content authors have been impacted by headless CMS and how Crafter CMS addresses these issues to provide authors with true headless CMS support for SPA applications.

Authors Forced into the Wayback Machine

Clumsy tools and processes for editing, reviewing and testing content offered by headless CMS platforms have content authors, the primary stakeholder of CMS platforms, wondering if they fell asleep and woke up back in the early 2000’s. Let’s take a look at some of the issues facing authors in the Headless CMS space.

Preview and In-Context Editing

Authors are accustomed to easy-to-use tools that enable them to visually locate, edit and publish content in websites or applications. Many headless CMS platforms provide no way for authors to preview their work prior to publishing it. Authors are forced to change content via content entry forms that are completely disconnected from the presentation.  In order to see what the content will look like on the website or in the application, they must publish the content. No author should be forced to work with slow and clumsy deployment iterations just to ensure their content fits and looks good on destination channels. That’s a major step backward.

Some CMS vendors and practitioners argue that by completely obscuring the presentation from the authors, content gets created in a fashion that is naturally more reusable. That’s not necessarily accurate. Content is more reusable when stored and transmitted in a fashion that is divorced from presentation — but that does not mean authors should be forced to ignore how their content will be presented on key channels. As the saying goes: “the medium is the message.” To most authors, the layout is just as important as the content itself. Authors need tools in their headless CMS that are on the level of the tooling they have become accustomed to with more traditional CMS platforms.

Delivering Editing and Preview tools for SPAs

Crafter CMS gives authors a full, interactive in-context preview and editing experience within SPAs. Authors have exactly the same tooling they have become accustomed to with traditional CMS.

Crafter’s preview capability can sit in front of any web-based application. The specific technology does not matter. Crafter’s preview and editing tools are front-end agnostic. One of the key reasons Crafter is able to do this is that Crafter CMS has always separated authoring and delivery into distinct but integrated components. We’ve been assuming a decoupled architecture for years.

Authors are the main users of any CMS and their needs must be properly addressed. A solid authoring experience with true in-context editing and preview is table-stakes for any headless CMS.

Personalization of Headless / API based Content

Another area that Headless CMS platforms have impacted authors is support for targeted/personalized content. Content authors need to be in control of the rules that drive the right content to the right people at the right place and time.

To support personalization your headless CMS must support the following two capabilities:

  • The ability to design, configure or code business rules behind the services.
  • The ability to preview content and functionality driven by personalization rules prior to publishing.

To date, most headless CMS platforms are simple repositories of content that are retrievable by API call. These systems offer no ability to create rules that can be executed based on the request parameter variations and characteristics of the requester.

Further, based on the previous section, we already know that most Headless CMS platforms don’t support pre-publishing preview of content — let alone the ability to preview content and experience as it would appear in different scenarios. There is little point in writing rules or variations of content without the means to properly test them prior to publishing.

Delivering Personalization for SPAs

Crafter CMS provides excellent support for targeted and personalized content even when content is delivered in a headless environment. Authors can configure scenarios and simple rules via the UI or leverage sophisticated rules built-in scripting languages by the development staff. Further, Crafter CMS provides rich preview tools that allow authors to impersonate different users and other types of content consumers within the authoring tools so that content and scenarios can be properly tested well before they are published.

The Need for Real Versioning and Branching

Most CMS platforms (traditional and headless) support some kind of basic versioning.  For the most part that support extends only to a single content item. Each item has its own history.  There is no way to look at the entire content collection at it existed at a specific point in time. That can be a major challenge for authors if they need to reference a previous campaign or support an audit.  The ability to maintain a version history is paramount and rudimentary versioning is typically not enough.

Another key feature that is missing from CMS platforms is the ability to support branches.  Branches allow us to work on different variants of the content at the same time. A branch not only allows concurrent work but it also enables us to sync up variants when it’s appropriate.  This extraordinarily important to authors who are supporting a re-branding or a major content overhaul. Most CMS systems can’t support that without content freezes and manual “double publishes”

Delivering Real Versioning and Branches

Crafter CMS’ repository is based on Git. Git provides authors using Crafter with a versioning system that is multi-file — meaning that an author can see the entire content collection at any moment in time in its history.  Git also provides support for branches which means that authors using Crafter can support difficult operational scenarios like re-brands without any content freezes or extra publishing.

Content freezes and Other Business Interruptions

New features, design refreshes, and integrations are a way of life for today’s customer-focused market. Agility and innovation at scale and speed are key to winning in this competitive landscape. Wherever you have innovation you will find developers and development process along with its procedures, tools, and environments. Businesses want innovation without disruption. Unfortunately with most CMS platforms, due to their architecture, business disruptions are exactly what content authors and other business users end up with. Outages and complicated deployment processes along with content freezes are realities for most innovative organizations use CMS.

When we look at the reason for the disruptions to the business activity we find that the development and testing process requires that we maintain independent testing environments and that we do code deployments. It’s during this maintenance activity that authors are often told to “hold off” or “stay out of the system.” Content freezes waste time and money and they hurt the business. SPAs and Headless CMS platforms do nothing to solve this problem.

In order to get content to lower environments like Dev, QA, and Test we have to take an export of Production. Authors need to stop working during exports. Imports on the lower environments are no fun either. Every environment needs to be painstakingly and individually updated. When coding is complete it has to be rolled out. Once again authors need to refrain from using the system. Complicated deployment procedures and models can take hours, days and in some cases even weeks. More valuable time and money are wasted.

Uninterrupted Business Activities with SPAs

Crafter CMS has an architecture that is different from other CMS platforms. Crafter is designed and implemented with repository technology specifically designed to handle these processes with minimal effort and friction. Crafter CMS leverages Git at its repository layer. In contrast to CMS platforms that rely on databases or JCR repositories that rely on exports and imports that require outages to sync data between environments, Crafter’s repository is meant to be distributed and as such it natively supports syncing between related stores. We call this push button process “Code Forward, Content Back”™ Crafter CMS allows developers along with their tools and process to live and work in harmony with content authors. This enables collaboration together without interruption and with much less friction so that a greater level of innovation and ongoing operations can all take place at the same time.


The net-net of it is simple. Regardless of any improvement in content reusability or ease of development, authors need to be able to do their jobs easily and efficiently without interruption. New tools and approaches need to move the ball forward, not back. Unfortunately for authors, headless CMS has lost major ground.

Some of these shortcomings can be overcome with time but all of the shortcomings we looked at here are architectural in nature. Architecture changes are the most difficult and time-consuming to make. You don’t have to wait for your current headless CMS to correct course. Crafter CMS is an open source headless CMS with the right architecture to take full advantage of the benefits of SPA applications provided to end users and developers while giving authors the tools, process and uninterrupted service they need to succeed.

What has your experience been working with SPAs and your CMS?

CMS for SPAs (1 of 4): Are Single Page Applications and Headless CMS a Slam Dunk?

From web-based applications like Gmail, Instagram, and Google Maps to websites like The New York Times and Facebook, real-world examples of Single Page Applications (SPA) are everywhere you look. Developers have been turning to SPA frameworks and architectures to create better, more usable, richer applications for their users.

Nearly every Content Management System (CMS) out there purports to support SPA frameworks and architectures. In this blog series, we’ll look at SPA use in the CMS space to learn more about what’s working, what’s not and what we can do about it. Let’s jump in.

SPAs are web-based applications built with rich UI frameworks that run client side and use partial page updates rather than full page refreshes to update data, content, and UI. SPAs have several advantages:

  • Great user experience: SPAs make the user experience more fluid with rich UI interactions that leverage asynchronous calls to the server for data and partial page refreshes.

  • Clean architecture: SPAs architecture promotes a clean separation of the frontend application and backend services and data. This approach makes development faster and more flexible.

  • Great frameworks: There are a lot of great frameworks that make developing SPAs faster and easier including React, Angular, Vue and others.

In the CMS space, SPA applications have been a win for developers but content authors have taken a major hit.

Developers have been keen to adopt SPA technology and CMS vendors have quickly responded with support for Headless CMS (also known as Content as a Service [CaaS]) support.

This shift has been highly promoted by CMS vendors because it has a number of advantages with respect to content management. Returning content as JSON or other presentation-less formats makes content more reusable within and across delivery channels. More importantly, by removing the consuming application from the CMS, the approach seems to solve one of the major challenges in the CMS space: development. It’s no secret that most developers have little interest in working within the toolsets and constraints of a CMS. Developers want control over their tools, frameworks, and process. CMS platforms are notoriously framework laden and don’t integrate well with DevOps tools and process.

The headless approach completely and neatly decouples development and content production workflows. On the surface, the headless approach appears to be a perfect evolutionary step for CMS. New delivery channels and integration opportunities show up all the time. In a customer experience driven market, anything that aids development and reduces time to market is seen as a winner.

However, the reality is that decoupling the CMS from the consuming application does not solve long-standing challenges such as refreshing content in lower environments — in fact, it complicates them. Worse, complete decoupling creates new problems. While software development and content production workflows are best divorced, their products: the code and content are not. There is a time/version specific relationship between the data model, the content and the code at any given moment. To illustrate the point: it’s often not possible to “point” an older version of an application at a newer version of the content and vice versa. It’s extremely difficult to troubleshoot or perform legal and regulatory audits without this capability. Yet, with the clear benefits of better products, a cleaner deployment architecture and faster, easier development in mind, the remaining issues are either unrecognized or a compromise many developers are willing to take. For developers, the SPA and headless approach have largely been a win.

The ironic and unfortunate “dirty little secret” in the CMS space of late is that the biggest losers of the headless CMS movement have been the content authors, CMS’s main stakeholder.

Today’s headless CMS options have the following drawbacks for authors:

  • There is no preview. Authors are used to in-context editing and preview. With headless, they are back to working in forms with no preview and are left to hope and a cumbersome process to ensure that the content will present properly.
  • There is no personalization. Marketers want to be able to target and personalize their messaging to the audience. Most headless CMS platforms don’t give them the capability to describe and execute targeting rules much less test and preview the various scenarios.
  • DevOps are still not supported: Despite investments in new technology, tools and process content authors see zero improvements in DevOps activities related to feature releases. Content freezes and outages are still commonplace during feature deployments with headless CMS.

It doesn’t have to be that way! With the right CMS architecture and platform content authors, developers and DevOps can work together on SPA based sites and applications at speed with the tools and processes they are accustomed to (like in-context editing) without stepping on each other’s toes.

A Better Approach to SPAs, CMS and DevOps

Crafter CMS is an open source CMS platform that takes a completely different approach to headless CMS, SPA development and DevOps. Crafter CMS has a very different architecture from traditional CMS platforms that are monolithic and rely on SQL databases and JCR repositories.

Crafter’s architecture is modular, flexible, and completely elastic with a repository and deployment design that is shared-nothing and based on Git. This architecture enables Crafter CMS the flexibility to support content authors, developers and DevOps to collaborate together to build and release new experiences with minimal friction while using the tools and workflows that work best for them.

Crafter CMS is an API first CMS that is:

  • UI agnostic and can support development and editing and preview for any UI framework
  • Capable of supporting rich, personalized headless CMS content
  • Git-based to support your development frameworks, tools, and process in a way that separates development and content production workflows without needing to separate the deployable (code and content) artifacts.


SPA applications combined with decoupled, headless CMS offer significant advantages in terms of end-user experience, development, and deployment. However, it’s clear that headless is not a silver bullet. Completely divorcing the CMS and the consuming application or site creates significant challenges for DevOps, support and other departments. More importantly, the SPA/headless pairing has left authors out in the cold by taking away the process and tools they have become accustomed to over the past decade.

These shortcomings can be addressed with a new breed of CMS architecture. What’s needed is a flexible headless CMS that support proper authoring tools and process regardless of front-end technology. Moreover, we need a CMS that embraces and integrates development tools and process rather than shunning them. That’s exactly what we’ve set out to do with Crafter CMS. Stay tuned for our next blogs in this series where we’ll dive deeper into these concerns and their solutions.

What has your experience been working with SPAs and your CMS?