Skip to main content

Part II: Agent Intelligence

Where specialized agents come alive: Security, Energy, and the testing infrastructure that proves they work.

The Agents Awaken​

With the foundation in place, it was time to build intelligent agents - specialized AI systems that could perceive, reason, and act within the building environment.

This is the story of:

  • Security Agent: Threat detection and coordinated response
  • Testing Infrastructure: Professional-grade validation
  • MCP & OPA Integration: Real tool execution and policy enforcement
  • Energy Agent: Mathematical optimization for efficiency

What You'll Learn​

Chapter 7: Security Agent​

LangGraph state machine for autonomous security responses.

Chapter 7.5: Testing Infrastructure​

2,300+ lines of tests, fixtures, mocks, and documentation.

Chapter 7.6: MCP & OPA Awakening​

Real MCP tool invocation and OPA policy evaluation (not mocks!).

Chapter 8: Energy Agent​

Scipy-based optimization for HVAC and demand response.

Agent Intelligence Complete ✅​

By the end of Part II, we had:

  • ✅ Security Agent with 5-state machine (MONITOR → ANALYZE → DECIDE → ACT → MONITOR)
  • ✅ 65+ comprehensive tests (unit, integration, E2E)
  • ✅ Real MCP tool execution (door locks, cameras, alerts)
  • ✅ Real OPA policy enforcement (production-ready)
  • ✅ Energy Agent with scipy optimization
  • ✅ Professional testing infrastructure

Continue the journey → Chapter 7: Security Agent