SKIMIX: Multi-Agent Harness-Time Scaling with Skill Mixture for Dynamic Harness Engineering
arXiv:2607.27994
The paper introduces SKIMIX, a multi-agent framework that enables AI agents to retrieve, combine, and evolve skills from a large library using embedding-based retrieval and submodular routing, improving performance on open-ended mathematical reasoning tasks while showing limited gains on multiple-choice benchmarks.
Abstract
AI agents increasingly rely on large skill libraries, but selecting, combining, and maintaining skills remains difficult. We propose SKIMIX, a multi-agent framework in which agents with different skill portfolios collaborate through iterative refinement. SKIMIX combines embedding-based skill retrieval, submodular anti-dilution routing, and adaptive skill evolution. Across six reasoning benchmarks, multi-agent collaboration substantially improves open-ended mathematical reasoning but offers limited or negative gains on multiple-choice tasks. Agent-count scaling is non-monotonic, and most improvements arise during the first refinement round. These results show that task characteristics determine whether skill-level ensembles help and provide practical guidance for scalable agent design.