OxyGent: Making Multi-Agent Systems Modular, Observable, and Evolvable via Oxy Abstraction

arXiv cs.AI / 4/29/2026

📰 NewsDeveloper Stack & InfrastructureSignals & Early TrendsModels & Research

Key Points

  • OxyGent is a new open-source framework aimed at making production-grade multi-agent systems easier to deploy in complex industrial settings by addressing scalability, observability, and autonomous evolution.
  • It introduces a unified “Oxy” abstraction that packages agents, tools, LLMs, and reasoning flows as pluggable atomic components to enable modular, Lego-like system composition.
  • For observability, OxyGent replaces fixed workflows with permission-driven dynamic planning that generates execution graphs at runtime, producing adaptive visualizations.
  • For continuous evolution, it adds “OxyBank,” an AI asset management engine that automates data backflow, annotation, and joint evolution of system components.
  • The authors report empirical evaluations and real-world case studies showing OxyGent as a robust, scalable foundation for multi-agent systems, and the project is released under Apache License 2.0.

Abstract

Deploying production-ready multi-agent systems (MAS) in complex industrial environments remains challenging due to limitations in scalability, observability, and autonomous evolution. We present OxyGent, an open-source framework driven by two core novelties: a unified Oxy abstraction and the OxyBank evolution engine. The unified abstraction encapsulates agents, tools, LLMs, and reasoning flows as pluggable atomic components, enabling Lego-like scalable system composition and non-intrusive monitoring. To enhance observability, OxyGent introduces permission-driven dynamic planning that replaces rigid workflows with execution graphs generated at runtime, providing adaptive visualizations. Furthermore, to support continuous evolution, OxyBank serves as an AI asset management platform that drives automated data backflow, annotation, and joint evolution. Empirical evaluations and real-world case studies show that OxyGent provides a robust and scalable foundation for MAS. OxyGent is fully open-sourced under the Apache License 2.0 at https://github.com/jd-opensource/OxyGent.