Built something you're proud of? Tell the story. A quick G2 review of App Engine or Build Agent helps other developers see what's possible on ServiceNow. Share your experience.

kristymerriam
Administrator

For developers with preferred IDE's and AI coding tools, ServiceNow's April Build Anywhere announcement offers maximum flexibility. Developers can build on ServiceNow using any front door. The ServiceNow SDK and agent skills now work across Claude Code, Cowork, OpenAI Codex, Cursor, Windsurf, and Antigravity. Work in the tools you know and love, then export to ServiceNow Studio like any other scoped app, with governance, security roles, and data models applied automatically. 

 

Before getting into the details, it’s worth seeing the full flow end to end. 

 

Watch: from prompt to production

 

 

As part of the SDK, ServiceNow has open-sourced specific build skills that enable coding agents like Claude Code to create native apps like a ServiceNow developer. It knows what ServiceNow platform objects to build, and how to build them using ServiceNow best practices. It understands the platform, the data model, the logic, and the scoped app structure that ServiceNow expects. Your coding agent writes in Fluent code directly into a sandbox on the ServiceNow platform, meaning the output is platform native by design. At no point do you need to open a ServiceNow instance in a browser. Everything is happening in the tools you know and love. 

 

kristymerriam_1-1775740126798.png

 

Where context comes into play 

Previously, an app that was first deployed had no awareness of what the platform already knew. (Org structure, data relationships, policies, integrations, etc.). Coming in Q2, the Context Engine will let apps inherit platform context such as org structure, policies, integrations once those integrations are configured. Your AI agents will start grounded instead of from scratch. Think of it like an organizational intelligence layer. The integrations already in place (Armis for asset graph, Pyramid Analytics for BI, data.world for data catalog) will give your app’s AI agents live enterprise context to work with from the very beginning. As you can see in this image, for any given request, the Context Engine takes into account things like policies, roles, and request history to make better decisions. 

 

kristymerriam_2-1775740144566.png

 

Free Build Agent calls just went up! 

  • Customer instances now get 100 free Build Agent calls, up from 25 
  • PDIs now get 25 free calls, up from 10 
  • Request a free PDI at developer.servicenow.com 

For anyone that has hit limits in the past, this offers more runway to work through your reasoning chains. One thing worth noting: descriptive prompts tend to get better results. The more context you give the agent about what you're building and why, such as the data model, the scope, or the business rule logic, the less back-and-forth you'll need and the more space you’ll have to experiment. Your platform expertise is still the variable that most determines the outcome. 

 

One more thing (worth knowing for internal conversations) 

AI, Workflow Data Fabric, and EmployeeWorks are now included by default in every ServiceNow level: Foundation, Advanced, and Prime. AI isn't only for the premium tier anymore. This is part of ServiceNow’s strategy to put the power of AI into the hands of every ServiceNow developer. 

 

What you need 

You only need two things before you start: 

  • Zurich or later (or a PDI, free at developer.servicenow.com) 
  • The Build Agent SDK installed. See what’s available and how to set it up: Start here (Available April 15)
  • New to the SDK or Claude Code? Earl Duque walks through the full setup step by step, from installation to deploying your first app. [Start the series here.]

The SDK and agent skills produce the same scoped app structure whether you're working out of your IDE or ServiceNow Studio. ATF tests and a readiness scan are generated either way. See what Build Agent can do. 

 

What to check once it deploys 

When an app deployed from an external tool hits your instance, it appears in ServiceNow Studio under your app list like any other scoped app. One thing to check before you do anything else- if your app uses AI agents, open AI Control Tower and confirm that they’re showing up. That your confirmation that observability is wired correctly. 

A few things to know: 

  • Branch management from external tools is currently limited. If you need to manage branches, use the Source Control menu or work directly in GitHub. Don’t try to drive that from your IDE yet. 
  • If deployment doesn’t complete cleanly, check that your credential record in ServiceNow matches the token you’re using in your external tool. Mismatched credentials are a common reason for a failed deployment. 
  • Full test deployment pipelines from external IDEs are still evolving. The SDK supports app creation and update flows well today. Automated pipeline management is coming, so keep an eye on the release notes. 

Your turn 

If you want a reason to start today, the #BuildMoreWithBuildAgent Challenge is live through May 4. Build something, share it, and you could win exclusive K26 swag. [See the challenge.]

 

12 Comments
aarondavidc
Tera Explorer

I’ve been experimenting with Cursor + ServiceNow’s new “Build Anywhere” approach and wanted to share something I’m building + get feedback from others on the platform.

I’m working on a concept called Hii (Human Impact Intelligence).

Problem:
In SOCs, help desks, and call centers—we track work, but not the human impact of that work in real time.

Result:

  • burnout
  • high turnover
  • repeated high-stress assignment patterns
  • errors under pressure

Most of this only becomes visible after the fact.

What I’m prototyping:
A ServiceNow-based approach that:

  • allows employees to request temporary work support (non-HR, operational)
  • translates that into assignment-aware rules (ex: reduce P1 exposure, adjust workload)
  • uses platform data to surface:
    • burnout risk
    • attrition patterns
    • performance risk under stress

One area I’m especially focused on:

Connecting HR-approved accommodations into operations
but only as non-PII operational rules

Example:

  • instead of exposing HR/medical details
  • the system only applies:
    • avoid P1 / critical incidents
    • limit concurrent workload
    • allow schedule flexibility

So accommodations can actually be enforced in assignment logic without exposing sensitive data.

Where this gets interesting:

Logging and learning from outcomes over time.

  • when work is rerouted or adjusted → log it
  • track downstream impact:
    • SLA performance
    • error/reopen rates
    • workload balance
    • retention patterns

The goal:
 Understand which operational supports actually reduce errors, burnout, and turnover
 and which don’t

So it becomes less about adding features—and more about
proving what improves outcomes and reduces cost.

End goal:
Feed these signals into assignment decisions (warn / block / reroute) in real time.

Would love input from others building in this space:

  • How are you handling the gap between HR accommodations and operational enforcement?
  • In high-volume environments, what patterns have you seen around manager overrides vs system protections?
  • For those using conversational AI (ex: MoveWorks), do you see a path for integrating this into the employee request/intake layer?

Using Cursor has made prototyping this much faster than expected.

Curious what others are building or seeing in this area.

 

Excited for the future of building!

Aaron

 

#Hii Human Impact Intelligence #futureofWork #ServiceNow HRSD ITSM #peopleFirst

 

SantoshK1857223
Mega Sage

The demo is mind blowing!!