Skip to content

References: AI Agent Architecture

  1. Intelligent Agent - Wikipedia - Defines agent architectures including perception-action loops, planning, memory, and tool use — directly foundational for this chapter's AI agent loop design, planning strategies, and the memory architecture that context graphs provide as persistent external memory.

  2. Artificial Intelligence: A Modern Approach - Wikipedia entry for Stuart Russell and Peter Norvig's textbook - Covers agent architectures, planning algorithms, and multi-agent systems at the depth needed to understand this chapter's agent loop patterns, ReAct, and Plan-and-Execute designs.

  3. Multi-Agent System - Wikipedia - Explains multi-agent system architectures including orchestration patterns, communication protocols, and coordination mechanisms — directly supporting this chapter's multi-agent orchestration section and agent authentication and authorization requirements.

  4. Artificial Intelligence: A Modern Approach (4th ed.) - Stuart Russell, Peter Norvig - Pearson - Chapters 2–5 cover rational agent architectures, planning, and search; Chapter 18 covers multi-agent systems — providing the foundational AI theory for this chapter's agent loop, planning, and orchestration patterns applied to context graph deployments.

  5. Hands-On Large Language Models - Jay Alammar, Maarten Grootendorst - O'Reilly Media - Chapters 13–15 cover LLM agent architectures, tool use, ReAct pattern, reflection, and agent evaluation — directly matching this chapter's treatment of ReAct, Plan-and-Execute, and Reflection agent patterns with context graph integration.

  6. Retrieval-Augmented Generation - Wikipedia - Covers RAG as the retrieval mechanism underlying agent read patterns — supporting this chapter's agent read pattern section and the context graph as persistent long-term memory that agents query before taking actions.

  7. Explainable Artificial Intelligence - Wikipedia - Covers XAI requirements for agent decision traceability including audit logs and decision justification — directly relevant to this chapter's agent trace, agent decision log, and graduated autonomy sections where transparency is the mechanism for building organizational trust.

  8. Access Control - Wikipedia - Explains access control models for software systems — supporting this chapter's agent authentication, agent authorization, and agent sandboxing sections that govern what context graph nodes and write operations each agent is permitted to access.

  9. Human-Computer Interaction - Wikipedia - Covers human-in-the-loop design patterns and feedback mechanisms — supporting this chapter's human-in-the-loop section and the graduated autonomy model where increasing agent trust is earned through demonstrated accuracy and transparency.

  10. Zero-Trust Security Model - Wikipedia - Defines the zero-trust security principle that no agent or system is inherently trusted — directly supporting this chapter's agent sandboxing, rate limiting, and the governance model where agents must prove trustworthiness through accumulated decision trace history before gaining expanded autonomy.