JDE Orchestration MCP Server
Expose existing JD Edwards Orchestrations to AI agents through typed MCP tools for discovery, introspection and execution.
Why Orchestrations are the right AI boundary
The safest way to let an AI agent do useful work in JD Edwards is not to give it direct database access or a generic automation shell. It is to expose the business actions your team has already curated as JDE Orchestrations.
An orchestration has a name, a purpose, a schema and security. That makes it a much cleaner unit of AI capability than an open-ended instruction such as “go change something in JDE”.
- Discovery lets the agent list only the orchestrations the configured JDE user is allowed to see.
- Introspection lets it inspect inputs and outputs before it calls anything.
- Execution runs through AIS under a real, traceable JDE user.
Where the AI MCP Server fits
The MCP server sits between the AI client and AIS. The agent sees named tools. JDE sees normal orchestration calls. Administrators keep the same security model and audit boundary they already understand.
This gives you practical agentic AI without turning the model into a privileged integration layer.
See where this fits in the product.
This supporting page explains one search intent in detail. The main AI MCP Servers page gives the full product overview, related use cases, licensing and enquiry path.
Back to AI MCP Servers