Agentic AI Reference Architecture

Orchestration · Software component

Termination Checker

Software componentOrchestrationOrchestration & Toolsarc:TerminationChecker

A control component that evaluates end conditions, such as reviewer approval or a maximum round count, after each message to decide whether a multi-agent conversation stops.

Responsibility. Decides after each message whether a multi-agent conversation has reached an end condition.

Also known as: Termination check function, End-condition validator

is invoked byis configured byreadsis invoked byConversational Agent Coordinator: is invoked byConversational Agent Coo…Iteration Limit Policy: is configured byIteration Limit PolicyShared Message History: readsShared Message HistoryIterative Refinement Coordinator: is invoked byIterative Refinement Coo…
Direct neighbourhood (hover for relationship types)

Relationships

is configured by structural

is invoked by dependency

reads dependency

Design guidance

Classification

Patterns
Bounded conversationApproval-based termination
Technologies
AutoGen
Quality attributes
Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)Cost efficiency
Risks mitigated
Unbounded agent conversations

Sources

  1. Ch2.4: T. Nguyen, "Multi-Agent Frameworks," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 2.4. ISBN: 9798244538229.
  2. Ch5.13: T. Nguyen, "Hybrid Decision Systems Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.13. ISBN: 9798244538229.