SwarmResearch: Orchestrating Coding Agents for Open-Ended Discovery

2026-07-02

AI Summary

Existing coding agents often converge prematurely to a single high-level strategy during extended operation, suffering from insufficient exploration. This work proposes a master-slave multi-agent architecture in which a Shepherd Agent dynamically guides multiple Search Agents based on global context, enabling parallel local search within independent Git branches to achieve diverse high-level exploration. By decoupling global guidance from local execution and integrating dynamic parallel scheduling with an evolutionary search mechanism, the approach overcomes the limitations of single-agent systems in context accumulation and state editing. Evaluated on 15 open-ended optimization tasks, the method matches or surpasses state-of-the-art performance on 13, demonstrating its effectiveness in enhancing high-level exploration and enabling adaptive parallel scalability.

Abstract

Long-running coding agents such as autoresearch can persistently discover optimizations for open-ended problems. However, they tend to converge onto a single high-level approach, then proceed with low-level edits while missing other superior approaches to the problem. We hypothesize two harness-level design choices contribute to this behavior: accumulating context in a single long-running agent and only exposing a single program state to edit. We introduce SwarmResearch, an orchestrator-subagent harness in which a Shepherd Agent uses global context to steer a population of Search Agents, each operating with local context in their respective git branch. On open-ended optimization tasks, SwarmResearch discovers better or comparable solutions to state-of-the-art LLM-guided evolution and multi-agent techniques on 13/15 tasks, driven by higher-level exploration. Compared with fixed scaling of serial and parallel agents, SwarmResearch's orchestrator-guided scaling discovers better-performing solutions by adapting parallelism at different search depths.

Problem

• coding agents
• open-ended discovery
• optimization
• multi-agent systems
• exploration

Innovation

• SwarmResearch
• multi-agent orchestration
• open-ended discovery
• adaptive parallelism
• Shepherd Agent