Bring your coding agent to governed AI development

With OutSystems Agent Experience, connect Cursor, Claude Code, Kiro and other coding agents to OutSystems via MCP. Build enterprise-grade apps using your favorite AI tools, with live system context, zero architectural drift and built-in safety.

null

Here are the 3 steps to get it done

Get Your Server Running

If you don't have a dedicated development environment, the OutSystems Personal Edition is your free forever sandbox to build apps and agents using the OutSystems Platform.

Start Here

Connect and use the MCP

Connect your AI coding agent to the OutSystems MCP and start using it. 

We launched MCP support for both ODC and O11.

Connect ODC MCP
Connect O11 MCP

Get Help and Connect

The hard part is done. 

Now join our dedicated Discord server to learn from others, share your story and keep doing cool things with Agent Experience. 

See you there!

Join our Discord

One governed lifecycle for any AI coding agent

OutSystems Agent Experience enables you to:

  • Cut Token Waste: Give agents instant, live system context, no re-learning required.
  • Prevent Architectural Drift: Enforce existing portfolio guardrails on every AI change.
  • Ship Real Code: Generate production-ready software mapped directly to your apps.

Loved by developers like you

John Issa
SVP of Technology

Ascent Developer Solutions

"Agent Experience helps our team keep pace with business demand while giving us better visibility into the underlying application logic. That's something we could never get from AI tools alone.

Thomas Eldrup
Solution Architect

NORMAL A/S

"The difference between Agent Experience and standalone AI tools is the enterprise governance and architectural guardrails built into the platform."

FAQs

Common questions about the OutSystems Agent Experience.


Looking for different information?


Join the Discord or 

Ask the Community → 

What is Agent Experience?

Agent Experience (AX) is the open platform layer that connects external AI coding tools to OutSystems enterprise context and governance. By exposing a standardized suite of APIs and MCP servers, AX allows AI tools to safely inspect system architecture, execute logic, and build applications on OutSystems without bypassing IT security or breaking existing software dependencies.

What is the Enterprise Context Graph (ECG)?

ECG provides a live context layer that gives AI understanding of your organization's software estate, data, and operations. It ensures every AI agent in your enterprise knows how your systems actually work: what connects to what, who's allowed to see it, and what changed. This way agents reason and act safely at every step of the software development lifecycle (SDLC) without creating hidden risk or requiring rework.

What does OutSystems use to connect to our existing AI coding tools?

OutSystems connects external AI coding tools using the Model Context Protocol (MCP), an open, widely adopted industry standard. Rather than forcing you into a proprietary ecosystem, MCP ensures your preferred AI coding tools can securely interact with the OutSystems platform.

How does building on OutSystems cut AI development costs and token spend?

Most AI coding tools guess blindly, wasting time and money figuring out your business setup. By giving AI direct access to how your systems work, it gets instant context from day one. This eliminates trial-and-error, cuts project costs, and delivers accurate results on the first pass.

How does OutSystems ensure AI agents follow enterprise security and governance policies?

All AI-driven code and actions pass through a single governed platform lifecycle. Agent Experience automatically applies role-based access controls (RBAC), security guardrails, human approval gates, and audit trails across every agentic workflow—preventing unauthorized bypass of IT policies before deployment.

Can developers use third-party AI tools like AWS Transform, Claude Code, or Cursor for legacy modernization projects?

Yes. Through standard MCP services, developers can connect their preferred tools directly into the platform. Agents access full system context to safely build, test, and deploy modernized components within a single governed lifecycle.