RSIHub: A Composable Framework for Agent-Evolution Methods

RSIHub is an open-source framework for building, comparing, and extending agent-evolution methods on shared infrastructure. Recipes compose reusable operators and declare mutable resources, while RSIHub supplies execution, evaluation, recovery, and lineage. The framework instantiates AHE, Hyperagents, A-Evolve, and GEPA without modifying the core, and evaluation is performed 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, though method rankings vary across settings and optimization gains do not always transfer to held-out tasks.