CoWork-X: Experience-Optimized Co-Evolution for Multi-Agent Collaboration System

TL;DR

CoWork-X introduces an active co-evolution framework for multi-agent collaboration that optimizes peer cooperation across episodes, reducing latency and token usage while improving performance in real-time tasks.

Contribution

The paper presents CoWork-X, a novel framework combining hierarchical skill retrieval and post-episode skill consolidation to enhance multi-agent collaboration under strict online constraints.

Findings

Achieves stable performance improvements in real-time collaboration benchmarks.

Reduces online latency and token consumption over time.

Demonstrates effectiveness in challenging Overcooked-AI-like tasks.

Abstract

Large language models are enabling language-conditioned agents in interactive environments, but highly cooperative tasks often impose two simultaneous constraints: sub-second real-time coordination and sustained multi-episode adaptation under a strict online token budget. Existing approaches either rely on frequent in-episode reasoning that induces latency and timing jitter, or deliver post-episode improvements through unstructured text that is difficult to compile into reliable low-cost execution. We propose CoWork-X, an active co-evolution framework that casts peer collaboration as a closed-loop optimization problem across episodes, inspired by fast--slow memory separation. CoWork-X instantiates a Skill-Agent that executes via HTN (hierarchical task network)-based skill retrieval from a structured, interpretable, and compositional skill library, and a post-episode Co-Optimizer that performs patch-style skill consolidation with explicit budget constraints and drift regularization. Experiments in challenging Overcooked-AI-like realtime collaboration benchmarks demonstrate that CoWork-X achieves stable, cumulative performance gains while steadily reducing online latency and token usage.