NEW
Give your coding agent the power of knowledge with GeneXus for Agents.
GeneXus for Agents is a set of skills, tools, and interfaces that enables AI agents to understand and operate on GeneXus Knowledge Bases with full context and control.
Rather than generating code from scratch, agents use knowledge to build and evolve systems that remain consistent, maintainable, and adaptable over time. GeneXus ensures deterministic, secure, and cost-effective generation behind every approved change. Compatible with any AI agent that supports the Model Context Protocol (MCP).
Knowledge Layer
What is GeneXus for Agents? The layer that makes your Knowledge Base accessible to AI agents.
Work in an integrated development environment (IDE) accessible from any device, with support for Mac, Windows, and Linux.
GeneXus for Agents enables AI agents to understand, generate, review, and propose changes to GeneXus objects with full platform context, while the GeneXus engine validates those changes and integrates them back into the development workflow.
In GeneXus, all business knowledge lives in the Knowledge Base - not in source code. GeneXus for Agents gives AI agents the ability to read, reason about, and contribute to that Knowledge Base, while GeneXus handles what it has always handled: generating deterministic, correct-by-construction code for any target technology with Symbolic AI.
The result: your team gets the speed of AI-assisted development without giving up the guarantees that make GeneXus fit for mission-critical systems.
In GeneXus, all business knowledge lives in the Knowledge Base - not in source code. GeneXus for Agents gives AI agents the ability to read, reason about, and contribute to that Knowledge Base, while GeneXus handles what it has always handled: generating deterministic, correct-by-construction code for any target technology with Symbolic AI.
The result: your team gets the speed of AI-assisted development without giving up the guarantees that make GeneXus fit for mission-critical systems.
How It Works
How does GeneXus for Agents work? A new way to build with GeneXus.
Work in an integrated development environment (IDE) accessible from any device, with support for Mac, Windows, and Linux.
GeneXus for Agents introduces a new way to interact with your Knowledge Base. Instead of working exclusively through the IDE, teams can now work from the IDE or from a CLI using AI agents powered by GeneXus Skills and the GeneXus MCP Server.
Knowledge Base objects are represented in text so agents can understand and modify them, while changes are validated and integrated back into the KB before generation. Agents propose changes. GeneXus validates and generates deterministically.
Knowledge Base objects are represented in text so agents can understand and modify them, while changes are validated and integrated back into the KB before generation. Agents propose changes. GeneXus validates and generates deterministically.
Architecture
Why not rely on prompt-based code generation alone?
Most AI coding tools generate code based on prompts. It may work at first, but it often lacks structure, consistency, and it fails in long-term maintainability.
With GeneXus for Agents, agents do not operate on isolated prompts. They work on your Knowledge Base - the source of truth for your system - while GeneXus delivers secure, cost-effective, and deterministic code generation, with built-in validation at every step. The result is faster iteration without giving up consistency, maintainability, or enterprise control.
With GeneXus for Agents, agents do not operate on isolated prompts. They work on your Knowledge Base - the source of truth for your system - while GeneXus delivers secure, cost-effective, and deterministic code generation, with built-in validation at every step. The result is faster iteration without giving up consistency, maintainability, or enterprise control.
Here’s how these approaches compare
|
|
Prompt-based AI coding | Traditional low-code | GeneXus for Agents |
|---|---|---|---|
| Source of truth | Prompts (ephemeral) | App definitions / visual models | Knowledge Base — persistent & technology-independent |
| Code generation | Probabilistic (LLM output) | Templates + drag-and-drop | Deterministic, correct by construction |
| Role of AI agents | Replace the developer | Not applicable | Contribute alongside the developer |
| System consistency | Manual review required | Platform-dependent | Enforced by the GeneXus engine |
| Technology evolution | Rewrite when stack changes | Platform-dependent | Automatic — new generator, same Knowledge Base |
| Enterprise readiness | Limited — requires heavy guardrails | Varies by vendor | Built-in validation, normalization, impact analysis |
What changes with GeneXus for Agents
What you gain
Work faster across tasks and modules
Use agents to explore, generate, review, and propose changes across modules and tasks with less friction and greater speed.
Extend what your system can do
Agents can assist with UI generation, integrations, auxiliary code, and repetitive tasks, while GeneXus keeps the core system consistent and deterministic.
Fit modern development workflows
Work with Git, branching, pull requests, code review, MCP, and CI/CD pipelines, integrating GeneXus into the workflows your team already uses.
Open by design, controlled by GeneXus
Use any LLM or agent compatible with MCP. No lock-in. Full control remains with the GeneXus engine.
Highly Scalable Applications
Build applications that scale effortlessly, regardless of user numbers or operational complexity.
Mission Critical Systems
Develop secure, resilient applications that guarantee uninterrupted operation of your organization’s critical processes, even in complex environments.
Core components of GeneXus for Agents
Label
GeneXus Skills: The specialization layer for AI agents.
GeneXus Skills define how agents understand and work with GeneXus objects, encapsulating patterns, naming conventions, object relationships, and best practices. This allows agents and coding CLIs to operate with consistency, speed, and precision.
GeneXus MCP Server: Standard access to your Knowledge Base
The GeneXus MCP Server implements the Model Context Protocol (MCP), enabling any compatible AI agent to interact with the Knowledge Base in a controlled way. Agents can read objects, query relationships, propose changes, and trigger compilation through a standardized interface.
Text Representation: Your Knowledge Base, readable by any LLM.
GeneXus for Agents represents Knowledge Base objects as structured text, allowing any LLM to understand, reason about, and modify them with full context.
Get started with GeneXus for Agents
Start exploring how AI agents can work on your Knowledge Base today.