Agentic Lybic: Multi-Agent Execution System with Tiered Reasoning and Orchestration

2025-09-13

AI Summary

Current desktop automation agents suffer significant performance degradation on complex, multi-step tasks due to weak coordination mechanisms and absence of quality control. This paper proposes a finite-state-machine (FSM)-based multi-agent architecture that decouples the system into four specialized roles—Controller, Manager, Executor, and Evaluator—to enable dynamic task scheduling, adaptive re-planning, and closed-loop quality assurance. We introduce a novel hierarchical FSM reasoning framework coupled with dynamic orchestration, integrating code-level operations, GUI interactions, and decision-theoretic analysis, while embedding continuous quality evaluation. Evaluated on the OSWorld benchmark, our system achieves a 57.07% success rate on 50-step tasks, establishing a new state-of-the-art and demonstrating superior generalization and robustness.

Technology Category

• Multiagent Systems: Multiagent Planning
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Humans and AI: Planning and Decision Support for Human-Machine Teams

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Autonomous agents for desktop automation struggle with complex multi-step tasks due to poor coordination and inadequate quality control. We introduce extsc{Agentic Lybic}, a novel multi-agent system where the entire architecture operates as a finite-state machine (FSM). This core innovation enables dynamic orchestration. Our system comprises four components: a Controller, a Manager, three Workers (Technician for code-based operations, Operator for GUI interactions, and Analyst for decision support), and an Evaluator. The critical mechanism is the FSM-based routing between these components, which provides flexibility and generalization by dynamically selecting the optimal execution strategy for each subtask. This principled orchestration, combined with robust quality gating, enables adaptive replanning and error recovery. Evaluated officially on the OSWorld benchmark, extsc{Agentic Lybic} achieves a state-of-the-art 57.07% success rate in 50 steps, substantially outperforming existing methods. Results demonstrate that principled multi-agent orchestration with continuous quality control provides superior reliability for generalized desktop automation in complex computing environments.

Problem

• Addresses poor coordination in multi-step desktop automation
• Solves inadequate quality control for autonomous agent systems
• Enables dynamic orchestration through finite-state machine architecture

Innovation

• Multi-agent system with finite-state machine architecture
• FSM-based routing for dynamic execution strategy selection
• Tiered reasoning with quality gating for error recovery