Blueprint Preview MCP design evaluation

Pega Blueprint Preview MCP is a design-time simulation capability available in Pega Blueprint. A connected AI-powered assistant can use the Blueprint application design as context during a conversation. The assistant can interact with exposed workflow tools and simulate how users can engage with the proposed application. 

Blueprint Preview MCP does not execute an implemented production workflow, create a production Case, or generate a Pega Platform™ application. Instead, it provides a simulation environment that helps teams evaluate and refine application designs before implementation. 

For step-by-step instructions on connecting an AI Assistant, see Connecting the Blueprint Preview MCP to Claude. 

Blueprint Preview MCP exposes workflow capabilities in a Blueprint design as MCP tools. A connected assistant can discover and invoke these tools during a conversation. The specific tools that are exposed depend on the Blueprint design and the configured workflows.  

The connected assistant uses the exposed workflow definitions and parameters as context for interactions with users. Delivery teams can use these interactions to evaluate how an AI assistant can guide users through the business processes in the Blueprint design. 

The simulation uses the Blueprint design instead of an implemented application. Therefore, it does not create production Cases or execute production workflows. 

Pega Blueprint Preview MCP helps you assess whether a workflow is suitable for AI agent interaction before implementation. Use Blueprint Preview MCP to identify ambiguous inputs, missing validation rules, exception paths, and actions that require human approval. This assessment helps you refine the workflow for agent-assisted interactions.

Blueprint Preview MCP provides a design-time simulation rather than a production solution. You still need to design and validate security, governance, data protection, auditing, error handling, and operational controls. Use the preview to refine the Blueprint and make informed implementation decisions rather than to validate production readiness.

Before you recommend an agent-assisted implementation, evaluate:

• Business suitability: Is agent interaction appropriate for this workflow and business outcome? 
• Tool boundaries: Does each exposed workflow represent a controlled business capability? 
• Input quality: Are required parameters clear, sufficient, and independently validated? 
• Case integrity: Does Pega retain control of Case state, validations, approvals, and business Rules? 
• Human oversight: Which consequential or irreversible actions require user confirmation or human review? 
• Security: What information is exposed, and how will identity and authorization be enforced in production? 
• Failure handling: How will ambiguity, invalid input, timeouts, retries, and duplicate requests be managed? 
• Auditability: Can the enterprise determine who requested an action, what the agent submitted, and what Pega executed? 
• Production gap: What additional runtime services and controls are required beyond the Blueprint simulation? 
• Portability: Is the workflow contract sufficiently clear to be used by different approved agent platforms? 

For an LSA, Blueprint Preview MCP is not about testing technology. The focus is validating whether an AI-powered interaction delivers the intended business outcome with appropriate safeguards, accuracy, and control. 

As a "mini-chief technology officer", an LSA uses Blueprint Preview MCP to identify design risks early, refine the workflow, and prepare the solution for implementation. The objective is not only to make an agent work but also to design an agent that operates responsibly and securely at enterprise scale. 

Check your knowledge with the following interaction: