Agentic AI Reference Architecture

Cognition · Software component

Plan Executor

Software componentCognitionCognition & Memoryarc:PlanExecutor

A component that iterates through planned steps, invoking tools or delegating to sub-agents for each, tracking progress and reporting failures.

Responsibility. Executes plan steps and tracks progress.

Also known as: Executor, HTN executor, Action executor

reads; writesdelegates toinvokesis orchestrated bytriggerswritesreceives data fromreadsreceives data fromis monitored byis configured byis orchestrated byis orchestrated byinvokesinvokesis overridden byis guarded byWorking Memory Buffer: reads; writesWorking Memory BufferWorker Agent: delegates toWorker AgentTool Executor: invokesTool ExecutorState-Graph Orchestrator: is orchestrated byState-Graph OrchestratorReplanner: triggersReplannerWorld Model State: writesWorld Model StateMCTS Planner: receives data fromMCTS PlannerExecution Plan: readsExecution PlanTree Search Controller: receives data fromTree Search ControllerPlan Deviation Monitor: is monitored byPlan Deviation MonitorAction Effect Model: is configured byAction Effect ModelPlan-and-Execute Controller: is orchestrated byPlan-and-Execute Control…Replanning Layer Arbiter: is orchestrated byReplanning Layer ArbiterParallel Tool Dispatcher: invokesParallel Tool DispatcherSemantic Function: invokesSemantic FunctionReflexive Safety Controller: is overridden byReflexive Safety Control…Plan Validator: is guarded byPlan Validator
Direct neighbourhood (hover for relationship types)

Relationships

is configured by structural

delegates to dependency

invokes dependency

reads dependency

writes dependency

receives data from dynamic

triggers dynamic

is guarded by control

is orchestrated by control

is overridden by control

is monitored by assurance

Design guidance

Quantitative guidance

As stated by the sources; verify before use.

Classification

Patterns
Sequential executionDAG-based parallel executionPlan-and-ExecuteParallel execution of unordered subtasks
Quality attributes
Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)Maintainability (ISO/IEC 25010)
Risks mitigated
Sequential bottleneck

Sources

  1. Ch1.2: T. Nguyen, "Core Agent Patterns," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.2. ISBN: 9798244538229.
  2. Ch1.5A: T. Nguyen, "Stateful Orchestration - Introduction and Core Concepts," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.5A. ISBN: 9798244538229.
  3. Ch1.6: T. Nguyen, "Stateful Orchestration - Pitfalls, Integration, and Synthesis," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.6. ISBN: 9798244538229.
  4. Ch5.2: T. Nguyen, "Tree-of-Thought (ToT) Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.2. ISBN: 9798244538229.
  5. Ch5.4: T. Nguyen, "Hierarchical Planning Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.4. ISBN: 9798244538229.
  6. Ch5.5: T. Nguyen, "Monte Carlo Tree Search Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.5. ISBN: 9798244538229.
  7. Ch5.6: T. Nguyen, "A* Search and Replaning," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.6. ISBN: 9798244538229.
  8. Ch8.2A: T. Nguyen, "Error Rates and Reliability," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 8.2A. ISBN: 9798244538229.