Persona-Aware Single-Instance Multi-Agent Coordination: The Hive Architecture

Zenodo (CERN European Organization for Nuclear Research) 2026 Preprint Open access

Read full text
Authors: Ltd. Maishi Shanghai Network Technology Co.
DOI: 10.5281/zenodo.18663706
Cited by: 0

Abstract

We present **Hive** (蜂巢), a persona-aware multi-agent coordination architecture that achieves the functional equivalent of $N$-agent parallel execution using a **single LLM instance** with persona switching. Unlike conventional multi-agent frameworks (e.g., Claude Agent Teams, AutoGen, CrewAI) that spawn $N$ independent model instances—incurring $O(N)$ token costs and $O(N)$ context isolation—Hive maintains a **unified context** and routes subtasks through a **PersonaEngine** that injects domain

Fields of Study & Subjects

Multi-Agent Systems and Negotiation Artificial Intelligence Applications Natural Language Processing Techniques

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
16 February 2026
Language
EN
Document type
Preprint