OpenEdge MCP Server leverages a containerized approach for deployment, which uses three distinct components. Understanding these components early helps you interpret the commands that follow.

Component overview

The following components are used for the OpenEdge MCP Server creation and deployment:
  • mcpgen (CLI utility)—A new utility designed for the containerized nature of the MCP Server, enabling administrative tasks, such as creating and configuring runtime instances.
    • Role: Serves as the profile assembler and lifecycle driver. It creates generated/<profile-_name> directories containing configuration files, certificates, and scripts and can also export production-ready bundles. For detailed information about the core CLI command set, refer to Use the mcpgen CLI for profile generation and management.
    • Typical action: Executes commands such as up, list, logs, export, remint-all, and client test calls.
    • Runs as: Local Python using the stdlib wheel through a wrapper script (.bat/.ps1, .sh)
  • MCP Server (runtime)—The MCP Server Docker image that delivers the core functionality of the OpenEdge MCP Server. It supports converting API specifications into MCP tools, managing authentication and authorization, tool scoping, routing requests, and calling downstream APIs.
    • Role: Serves generated MCP tools from your OpenAPI specification, including authentication, scopes, and guardrails.
    • Typical action: Handles client tool discovery and invocation.
    • Runs as: openedge-mcp-server container image
  • Utilities image—A Docker image that assists mcpgen to create profiles, generate certificates, create service account JWT tokens, and make MCP Client calls.
    • Role: Provides an internal helper container invoked by mcpgen for operational tasks such as service account key generation, token/JWT creation, and certificate handling. Not intended for direct standalone use
    • Typical action: Performs key and credential bootstrap and diagnostics, which are triggered automatically
    • Runs as: openedge-mcp-utils container image

Rationale for component separation

The components are separated to ensure security, efficiency, and streamlined onboarding for different deployment stages. The following points explain why each component has its own role:
  • Lean production image—The server ships only request-path code for efficiency.
  • Secure separation—Credential and key minting occurs in a dedicated utilities image, kept out of the runtime environment, and orchestrated by mcpgen.
  • Fast onboarding—mcpgen composes all elements deterministically without requiring heavy dependency installations.
Note: Due to the powerful and administrative nature of mcpgen, it is not recommended to deploy this utility in any production environment running an MCP Server. The export process exists to prepare an MCP Server runtime instance for self-contained operation, eliminating the need for additional tooling.

How the components work together

Understanding how the components interact helps you visualize the overall workflow. This model shows the sequence of actions from development to production:
  1. Use the mcpgen CLI to generate a profile, manage configuration, and orchestrate lifecycle operations for OpenEdge MCP environments. During profile generation process, mcpgen launches the MCP Server and automatically invokes the utilities image whenever privileged operations, such as key generation or certificate handling are required.
  2. Protect the execution of the MCP utilities. All operations involving credentials, certificates, or diagnostics are wrapped by mcpgen commands for security and simplicity.
  3. In production, only the MCP Server container runs continuously to serve MCP tools. The mcpgen CLI and utilities image are used occasionally for maintenance tasks such as key rotation or configuration updates, typically from a controlled workstation or CI/CD pipeline.