Hey!
Did any one here try out the Pega PAI’s integration MCP server for reporting?
Recently we had a use case where business had a requirement get any field available in Peag.
So thought of trying out this approach for the same, and found it fascinating that what can be done with this.
I am looking for any suggesting which can allow achieving dynamic data modeling, something like independent of the specific project and can get all the columns seamlessly.
Use one metadata-driven reporting API in Pega, not a project-specific MCP prompt. Define a standard Data View or application-specific REST API that exposes only approved report fields, then let the MCP server first read the metadata endpoint and build the column list from that instead of guessing from case structure. Pega’s DX API supports metadata-style access for data objects and data views, including queryable fields, and Pega can publish application-specific REST APIs with OpenAPI documentation for the same contract. This gives you dynamic reporting that works across projects, but stays governed because every field is explicitly modeled and exposed by Pega. It also avoids the problem of trying to pull every property in the application, which becomes inconsistent and unsafe across different case types and implementations. Since Pega now supports MCP in its agent interoperability direction, MCP should sit on top of this governed API layer, not replace the reporting model itself.