We introduce a modular harness design for LLM agents that composes of perception, memory, and reasoning components, enabling a single LLM or VLM backbone to tackle a wide spectrum of multi turn gaming environments without domain-specific engineering. Using classic and modern game suites as low-barrier, high-diversity testbeds, our framework provides a unified workflow for analyzing how each module affects performance across dynamic interactive settings. Extensive experiments demonstrate that the harness lifts gameplay performance consistently over un-harnessed baselines and reveals distinct contribution patterns, for example, memory dominates in long-horizon puzzles while perception is critical in vision noisy arcades. These findings highlight the effectiveness of our modular harness design in advancing general-purpose agent, given the familiarity and ubiquity of games in everyday human experience.
@article{arxiv.2507.11633,
title = {General Modular Harness for LLM Agents in Multi-Turn Gaming Environments},
author = {Yuxuan Zhang and Haoyang Yu and Lanxiang Hu and Haojian Jin and Hao Zhang},
journal= {arXiv preprint arXiv:2507.11633},
year = {2025}
}