MCP is a standardized mechanism that allows AI systems to dynamically access and invoke external tools. When Blueprint exposes its capabilities through MCP, any MCP-compatible AI - such as IBM Project BOB - can discover, select, and execute Blueprint services autonomously. This enables truly agentic behavior, where the AI determines which tools to use, in which order, and for what purpose, based on the user’s intent.

Deep IBM i Insight Through Agentic AI

Blueprint maintains a complete structural model of an IBM i application landscape, represented as a node-and-edge graph. This model includes:

• Programs

• Tables and physical files

• Logical files and views

• Stored procedures

• Triggers and dependencies

Unlike traditional ER diagrams, Blueprint captures the entire operational relationship between components, including indirect updates and runtime dependencies.

Through MCP, this rich model becomes accessible to Project BOB in a machine-readable format. Blueprint services return the same authoritative data used in the Blueprint UI, but expressed as structured text optimized for AI consumption.

Practical Examples with Project BOB

Program usage analysis
If a user asks Project BOB a question such as:

“What does the stored procedure SKANOL_BATCH call?”

Project BOB evaluates the available MCP tools and automatically selects Blueprint’s get_used_programs service. Blueprint retrieves dependency data directly from the IBM i system and returns a detailed call tree, allowing BOB to present accurate results without manual investigation.

Field usage search
A question like:

“Where is the field OTSTAT used?”

Triggers Project BOB to invoke Blueprint’s search_source service. This scans source code and database definitions to identify all programs, tables, and views that reference the field.

Change Impact Analysis at Scale

The true strength of the agentic approach becomes evident with multi-step questions, such as:

“Which programs need to be recompiled if the size of OTSTAT changes?”

In this case, Project BOB autonomously orchestrates multiple Blueprint services:

1. It first calls search_source to identify all references to the OTSTAT field.

2. It then calls get_used_programs to build a full dependency tree.

3. The result is a complete, accurate change-impact analysis across the system.

Tasks that previously required days or even months of manual analysis can now be completed in minutes. Project BOB can even proceed to initiate recompilation workflows based on the results.

Blueprint Intelligence Beyond MCP

In addition to its MCP-exposed services, Blueprint includes its own AI capabilities. These automatically analyze and document code structure and execution flow, producing clear, human-readable explanations that are visible directly in the Blueprint user interface. This contextual documentation further enhances the value of the data consumed by Project BOB.

Summary

By combining Blueprint’s comprehensive IBM i application model with MCP and IBM Project BOB’s agentic AI, organizations gain:

• Accurate, real-time dependency insight

• Automated impact analysis

• Faster and safer change management

• Reduced manual effort and risk

This integration demonstrates how Blueprint is not only compatible with IBM Project BOB, but also purpose-built to unlock its full potential.