Skip to Content
ConceptsAgent architectures (comparison)

Agent architectures (comparison)

This page compares cutting-edge agent architectures on a Level 1 to 4 spectrum and explains where Agentron sits: it uses all four levels—Level 1 (chat/RAG), Level 2 (workflows), Level 3 (heap planner routing), Level 4 (heap, create tools/agents on the fly)—with cost control and debuggable topology.


Taxonomy: design-time vs runtime structure

Agent architectures sit on a spectrum from design-time (human-controlled) to runtime (model-controlled) structure:

LevelArchitectureWho decides structureUsed in productionAgentronNotes
1ReAct + tools (orchestrator)Human picks tools; model picks order✅ Yes (ChatGPT, Claude, Gemini)YesSingle agent, fixed tool set, dynamic tool selection/ordering. Thought to Action to Observation; cost control and loop limits are mandatory. Agentron uses this for chat and RAG (notebook and inline): one ReAct agent, fixed tool set, model picks order.
2Multi-agent, static topologyHuman designs graph and roles✅ Yes (LangGraph, CrewAI, DSPy)YesPredefined roles and flows. Predictable and debuggable; dominant in production. Agentron uses this for workflows: human-designed graph, configurable rounds.
3Multi-agent, dynamic routingModel chooses which agents to invoke⚠️ EmergingYesModel decides routing at runtime; more flexible, harder to debug. Agentron uses this in heap: the planner (LLM) chooses which specialists to invoke and in what order per request—model-chosen routing over specialists.
4Recursive / self-building (RLM)Model builds its own architecture🔬 Frontier researchYesModel writes agents/structure on the fly; less production-ready, high variance. Agentron uses this in heap: the system can create new tools and agents on the fly; structure and novel agents are model-determined at runtime.

Where Agentron fits: Agentron uses all four levels of the taxonomy. Level 1 (ReAct + tools): chat and RAG, notebook and inline block chat (retrieve, cite, optional web search); single agent, model picks tool order. Level 2 (static topology): workflows, human-designed graph, configurable rounds. Level 3 (dynamic routing): heap, the planner chooses which specialists to invoke and in what order per request (model-chosen routing). Level 4 (recursive / self-building): heap, the system can create new tools and agents on the fly; structure and novel agents are model-determined at runtime. The runtime builds a DAG from the plan and runs it. Specialist definitions can be human-defined or model-created.

Reference: The design-time vs runtime taxonomy is widely discussed—e.g. A Taxonomy of Agent Architectures: ReAct to Recursive Language Models  (Gordon, 2026) and the AI Agent Design Patterns catalog . For recursive/model-built architectures see the MIT Recursive Language Models  paper (Zhang, Kraska, Khattab).


Orchestration patterns (multi-agent)

When you use multiple agents, orchestration pattern matters:

PatternDescriptionProsCons
Hierarchical (supervisor/worker)Central supervisor decomposes and delegatesGood for compliance, clear ownershipRisk of bottleneck
Sequential pipelineLinear handoff (e.g. research to analyze to write)Good for document/review flowsFragile to single failure
Parallel (ensemble)Same task in parallel; aggregate resultsGood for brainstorming, diversityHigher cost, coordination
Event-driven (pub/sub)Message broker; agents react to eventsGood for high volume, real-timeMore moving parts
Peer-to-peer (mesh)No central coordinatorResilientHarder to debug

Where Agentron fits: Agentron uses the hierarchical (supervisor/worker) pattern for heap: the planner delegates to specialists. That pattern can implement the others for agents (e.g. the supervisor can delegate in sequence, in parallel, or via events), so one orchestration style covers varied workflows. Internally, event-driven patterns are used for workflow execution, chat pub/sub, and the run queue; see Event-driven architecture.

Agentron: Without heap, chat and RAG use a single ReAct agent with tools. With heap, the planner outputs a plan (priorityOrder) and the runtime builds a DAG from it, running specialists by level (e.g. general, workflow, tools, knowledge). Heap is hierarchical with a model-assembled DAG per request and can create new tools and agents on the fly; workflows stay human-defined static topology.


Production and safety (2026)

Cutting-edge agentic systems need explicit controls:

ConcernGuidance
Cost and loopsUnbounded ReAct loops cause runaway cost. Use max steps, loop detection, and tool-call budget.
Tool reliabilityFlaky tools trigger retries and cost spikes. Tools should be well-specified and optionally guarded (timeouts, circuit breakers).
HandoffsIf multi-agent, treat handoffs as versioned contracts; validate at boundaries; transfer full context or clear summary.
Failure modesHallucination propagation, cascading failures, coordination failures. Use bulkhead isolation, timeouts, graceful degradation.
Provider-agnostic layerPrefer one abstraction over provider-specific SDKs; support multiple LLMs and optional MCP for tools.

Agentron implements max steps, bounded rounds in workflows, and a provider-agnostic runtime (OpenAI, Anthropic, Ollama, remote LLMs, MCP tools).


System-theoretic view (2026)

Agentic systems can be decomposed into:

SubsystemRole
Reasoning & world modelHow the agent reasons (e.g. LLM).
Perception & groundingHow the agent gets real-world signal (e.g. RAG, search).
Action executionHow the agent acts (tools, code, APIs).
Learning & adaptationOnline learning, feedback loops (optional in v1).
Inter-agent communicationHandoffs, contracts (relevant when multi-agent).

Agentron (v1): Strong on perception (retrieval, RAG) and action (tools, workflows); reasoning is the LLM. Learning and adaptation: auto-mode learning for created agents is supported; user-guided learning is possible for both the Agentron agent and created agents. Deep inter-agent protocols can be deferred until needed.


Where Agentron fits (summary)

  • Chat / RAG (no heap): Level 1 (ReAct + tools). Retrieve, cite, optional web search; max steps and cost control.
  • Workflows: Level 2 (static topology). Human-designed graph, configurable rounds.
  • Heap (chat): Level 3 (dynamic routing)—planner chooses which specialists and in what order per request—and Level 4 (recursive / self-building)—system can create new tools and novel agents/architectures on the fly. The runtime assembles a DAG from the plan and runs it. Specialists include general (Q&A, web search), workflow, agent, tools, knowledge (connectors), improve_run, improve_heap, improve_agents_workflows; execution DAG and new tools/agents are model-built at runtime.
  • No unbounded ReAct loops; yes to provider-agnostic LLMs and MCP.

Research & further reading

For in-repo deployment and runtime boundaries, see the architecture  doc in the repository.

Last updated on