This topic outlines the process for setting up the OpenEdge MCP Server in development environments. The following steps represent high-level actions in the workflow. These steps ensure your MCP Server is properly configured in the development environment, before you enforce security best practices and prepare for deployment to production. For detailed instructions and configuration options, refer to the respective linked topics provided in each step:
  1. Step 1: Download and unpack the distribution package.
  2. Step 2: Run the container image load script.
  3. Step 3: Verify your environment.
  4. Step 4: Provide OpenAPI specification.

  5. Step 5: Generate and start a profile.

  6. Step 6: Validate the tool catalog.
What's next: After completing the installation and initial setup in your development environment, start the MCP Server with the generated profile and test it by sending a natural language request mapped to API operations. To do this, use a client that implements the Model Context Protocol (MCP) and has access to large language models (LLMs). For example, Visual Studio Code (VSCode) provides integrated MCP Client support and enables interaction with LLMs for natural language query processing to MCP tools.

Further, it is essential to enforce security best practices and harden your profile before deploying to production. This includes enabling TLS or mTLS, configuring authentication and authorization, rotating keys and tokens, vaulting secrets, and applying runtime guardrails. For more information on the security and pre-production hardening checklist, see Enforce security best practices before deployment.

Only after you have followed each step in the checklist should you export your hardened profile and proceed with production deployment. Following these measures help ensure your OpenEdge MCP Server is secure, compliant, and production-ready. For more information on deploying OpenEdge MCP Server to production, see Deploy OpenEdge MCP Server to production.