MEGA Hub

RSIHub: A Composable Framework for Agent-Evolution Methods

Authors

Do you know Zimu Wang?You can claim authorship or link another user.Do you know Xiaobo Wang?You can claim authorship or link another user.Do you know Siyu Ye?You can claim authorship or link another user.Do you know Jiaqi Li?You can claim authorship or link another user.Do you know Silin Chen?You can claim authorship or link another user.Do you know Yuling Shi?You can claim authorship or link another user.Do you know Ruobing Wang?You can claim authorship or link another user.Do you know Jingjing Zhang?You can claim authorship or link another user.Do you know Chaofan Wang?You can claim authorship or link another user.Do you know Kai Wu?You can claim authorship or link another user.Do you know Mengnan Qi?You can claim authorship or link another user.Do you know Kai Cai?You can claim authorship or link another user.Do you know Xiaodong Gu?You can claim authorship or link another user.Do you know Zilong Zheng?You can claim authorship or link another user.

Abstract

Agent-evolution methods use execution feedback to improve prompts, skills, tools, and orchestration, but their implementations often rebuild the same experimental machinery. We present RSIHub, an open-source framework for building, comparing, and extending these methods on shared infrastructure. Recipes compose reusable operators and declare mutable resources; RSIHub supplies execution, evaluation, recovery, and lineage. Researchers can focus on evolution policy while evaluation stays outside the mutable surface and results remain tied to candidates and evaluation conditions. We instantiate AHE, Hyperagents, A-Evolve, and GEPA without modifying the framework core, keeping method-specific logic in recipes and operators. We evaluate them with MiniSWE and Codex on Terminal Bench 2 and τ^3 Banking under fixed model and evaluator configurations. The largest observed overall pass@1 gain is 13.5 percentage points, while method rankings vary across settings and optimization gains do not always transfer to held-out tasks. RSIHub provides a common protocol for comparing evolution policies, inspectable candidate histories, and reusable components for developing new methods.

Community

00