@Jason Seoud 

May be this link will give you some idea.

https://www.linkedin.com/pulse/bridging-pega-claude-introducing-mcp-server-venkata-pardha-saradi-hzpoc

But this is limited to what you define in the MCP. It cannot create rules in Pega (at least not yet I guess but its just a matter of time :) ) but can certainly create or fetch cases using defined APIs

@Jason Seoud Pega does not natively support Model Context Protocol (MCP) as an out-of-the-box integration method. MCP is typically used in AI/ML environments for sharing models and context information between agents or services, especially in collaborative AI systems. Pega does support integration with external AI/ML models using REST APIs, Web Services, or Kafka streams. If the client’s AI agents are using MCP, you would need to build a custom middleware or adapter that translates MCP messages into RESTful or other supported formats that Pega can consume. This allows Pega to interact with external agents while maintaining its security and integration standards. So while Pega doesn't directly support MCP, interoperability can still be achieved with custom integration.

@Jason Seoud In the context of MCP, pega can come into play by exposing APIs as tools in MCP. So if there is a MCP server setup with bunch of tools, one or more of those tools can be Pega application APIs. 

The incoming prompt can then make use of those tools to interact with Pega applications data(cases, other integrations etc.), get the response and give it to MCP server which will then be transformed into output to the prompt. 

@Jason Seoud  Pega supports emerging agent interoperability standards including Model Context Protocol (MCP) and Agent-to-Agent (A2A). This is supported in 25.1 version. We have the ability to create the MCP connector from the dev Studio.

https://docs.pega.com/bundle/platform/page/platform/gen-ai/connecting-agent-mcp-protocol.html