Can AI Scientists Coordinate at Runtime?

2026-09-30

AI Summary

Existing AI scientist systems are constrained by fixed workflows, limiting dynamic coordination at runtime. To address this, this work proposes the RAC framework, which introduces a novel runtime agent orchestration mechanism. By leveraging execution-time dynamic selection, scoped contract allocation, and artifact-based verification, RAC enables non-blocking collaboration that emulates the flexible division of labor characteristic of human scientists. Evaluated on the ResearchClawBench benchmark, the proposed runtime selection mechanism significantly improves system performance, demonstrating its effectiveness in multi-agent scientific research automation. However, the practical gains derived from complex coordination mechanisms remain bounded by computational budgets.

Abstract

Multi-agent AI scientists have shown improving performance across a diverse range of tasks. Yet a common approach is design-time agentic orchestration, which typically relies on fixed workflows. In contrast, human scientists coordinate and adjust their division of labor at runtime. We therefore ask: can AI scientists also coordinate at runtime? To this end, we introduce Runtime Agent Coordination (RAC), which selects agents from existing AI-scientist hosts during execution, assigns scoped work contracts, and provides artifact-grounded verification. Verification informs subsequent agents without blocking transitions or discarding artifacts. We conduct a single-seed exploratory evaluation across Agent Laboratory, EvoScientist, and ARK on ResearchClawBench, preserving host models, tools, and permissions under host-calibrated budgets. Four cumulative conditions separate native execution, runtime communication, runtime selection, and the combined addition of contracts and verification. Runtime selection yields the highest observed mean score for each host; adding contracts and verification reduces these means, with host-dependent outcomes relative to native execution. These results motivate runtime coordination while exposing the limits of additional coordination mechanisms under constrained budgets. Code is available at https://github.com/systemind-team/Runtime-AI-Scientist.

Problem

• Multi-agent AI scientists
• Runtime coordination
• Agentic orchestration
• Division of labor

Innovation

• Runtime Agent Coordination
• Multi-agent AI Scientists
• Dynamic Agent Selection
• Work Contracts
• Artifact-grounded Verification