From Agent Loops to Structured Graphs:A Scheduler-Theoretic Framework for LLM Agent Execution
Hu Wei
Why It Matters
What makes this one worth your time
This paper matters because it addresses critical weaknesses in the current LLM agent execution paradigm, potentially leading to more reliable and debuggable AI systems.
A new framework proposes structured graphs for more controllable LLM agent execution.
Summary
The paper introduces a novel framework, Structured Graph Harness (SGH), to address the limitations of the Agent Loop paradigm in LLM-based agent execution by applying scheduler-theoretic principles, offering improved controllability and verifiability through a static DAG representation of control flow.
Key contributions
- A scheduler-theoretic framework for LLM agent execution that applies classical scheduling theory to address challenges in current paradigms.
Notable insights
- The use of a static DAG for control flow can significantly enhance the verifiability and controllability of LLM-based agents.
Possible limitations
- The paper is a position paper and design proposal without empirical results or a production implementation.
Abstract
arXiv:2604.11378v1 Announce Type: new Abstract: The dominant paradigm for building LLM based agents is the Agent Loop, an iterative cycle where a single language model decides what to do next by reading an ever growing context window. This paradigm has three structural weaknesses: implicit dependencies between steps, unbounded recovery loops, and mutable execution history that complicates debugging. We characterize the Agent Loop as a single ready unit scheduler: at any moment, at most one executable unit is active, and the choice of which unit to activate comes from opaque LLM inference rather than an inspectable policy. This perspective places Agent Loops and graph based execution engines on a single semantic continuum. We propose SGH, Structured Graph Harness, which lifts control flow from implicit context into an explicit static DAG. SGH makes three commitments: execution plans are immutable within a plan version, planning execution and recovery are separated into three layers, and recovery follows a strict escalation protocol. These choices trade some expressiveness for controllability, verifiability, and implementability. Our contributions are fourfold: a scheduler unified framework that applies classical scheduling theory to LLM agent execution and identifies challenges introduced by non deterministic LLM nodes; a trade off analysis of controllability, expressiveness, and implementability across 70 surveyed systems; a formal specification including a node state machine with termination and soundness guarantees; and an attributable experimental framework with a seven group design for future validation. This is a position paper and design proposal. We provide a theoretical framework, design analysis, and experimental protocol, not a production implementation or empirical results.