Zhiyao Cui

Hao Li, Chunjiang Mu, Jianhao Chen et al.

The rapid proliferation of Claude agent skills has raised the central question of how to effectively leverage, manage, and scale the agent skill ecosystem. In this paper, we propose AgentSkillOS, the first principled framework for skill selection, orchestration, and ecosystem-level management. AgentSkillOS comprises two stages: (i) Manage Skills, which organizes skills into a capability tree via node-level recursive categorization for efficient discovery; and (ii) Solve Tasks, which retrieves, orchestrates, and executes multiple skills through DAG-based pipelines. To evaluate the agent's ability to invoke skills, we construct a benchmark of 30 artifact-rich tasks across five categories: data computation, document creation, motion video, visual design, and web interaction. We assess the quality of task outputs using LLM-based pairwise evaluation, and the results are aggregated via a Bradley-Terry model to produce unified quality scores. Experiments across three skill ecosystem scales (200 to 200K skills) show that tree-based retrieval effectively approximates oracle skill selection, and that DAG-based orchestration substantially outperforms native flat invocation even when given the identical skill set. Our findings confirm that structured composition is the key to unlocking skill potential. Our GitHub repository is available at:https://github.com/ynulihao/AgentSkillOS.

Siyue Ren, Zhiyao Cui, Ruiqi Song et al.

Social norms play a crucial role in guiding agents towards understanding and adhering to standards of behavior, thus reducing social conflicts within multi-agent systems (MASs). However, current LLM-based (or generative) MASs lack the capability to be normative. In this paper, we propose a novel architecture, named CRSEC, to empower the emergence of social norms within generative MASs. Our architecture consists of four modules: Creation & Representation, Spreading, Evaluation, and Compliance. This addresses several important aspects of the emergent processes all in one: (i) where social norms come from, (ii) how they are formally represented, (iii) how they spread through agents' communications and observations, (iv) how they are examined with a sanity check and synthesized in the long term, and (v) how they are incorporated into agents' planning and actions. Our experiments deployed in the Smallville sandbox game environment demonstrate the capability of our architecture to establish social norms and reduce social conflicts within generative MASs. The positive outcomes of our human evaluation, conducted with 30 evaluators, further affirm the effectiveness of our approach. Our project can be accessed via the following link: https://github.com/sxswz213/CRSEC.

Zhiyao Cui, Chenxu Wang, Shuyue Hu et al.

Producing presentation slides automatically entails coordinating narrative structure with page-level graphic design under strict spatial constraints. For such structured multimodal tasks, a well-organized design process is essential to ensure the final quality of slides. Existing approaches rely on fixed templates or directly emit executable code, thereby both limiting the creative layout-design capabilities of LLMs and bypassing the essential slide-page design step. To address these limitations, this paper (1) proposes a hierarchical slides generation workflow, DeepSlides, that systematically organizes slide design tasks without any predefined template or style, decoupling slide-page design from implementation; (2) introduces SlideDesign, a dataset tailored specifically for slides generation tasks; and (3) presents a multi-agent reinforcement learning training paradigm and trains a couple of models, SlideQwens, for slide design and implementation. Experimental results demonstrate that our proposed framework outperforms baseline methods on evaluated metrics and achieves superior performance in human preference evaluations. The dataset and code are available at https://github.com/sxswz213/DeepSlides.

Xin Wang, Zhiyao Cui, Hao Li et al.

Vision language model (VLM)-based mobile agents show great potential for assisting users in performing instruction-driven tasks. However, these agents typically struggle with personalized instructions -- those containing ambiguous, user-specific context -- a challenge that has been largely overlooked in previous research. In this paper, we define personalized instructions and introduce PerInstruct, a novel human-annotated dataset covering diverse personalized instructions across various mobile scenarios. Furthermore, given the limited personalization capabilities of existing mobile agents, we propose PerPilot, a plug-and-play framework powered by large language models (LLMs) that enables mobile agents to autonomously perceive, understand, and execute personalized user instructions. PerPilot identifies personalized elements and autonomously completes instructions via two complementary approaches: memory-based retrieval and reasoning-based exploration. Experimental results demonstrate that PerPilot effectively handles personalized tasks with minimal user intervention and progressively improves its performance with continued use, underscoring the importance of personalization-aware reasoning for next-generation mobile agents. The dataset and code are available at: https://github.com/xinwang-nwpu/PerPilot

Hangfan Zhang, Zhiyao Cui, Jianhao Chen et al.

Multi-agent debate (MAD) has gained significant attention as a promising line of research to improve the factual accuracy and reasoning capabilities of large language models (LLMs). Despite its conceptual appeal, current MAD research suffers from critical limitations in evaluation practices, including limited benchmark coverage, weak baseline comparisons, and inconsistent setups. This paper presents a systematic evaluation of 5 representative MAD methods across 9 benchmarks using 4 foundational models. Surprisingly, our findings reveal that MAD often fail to outperform simple single-agent baselines such as Chain-of-Thought and Self-Consistency, even when consuming significantly more inference-time computation. To advance MAD research, we further explore the role of model heterogeneity and find it as a universal antidote to consistently improve current MAD frameworks. Based on our findings, we argue that the field must stop overvaluing MAD in its current form; for true advancement, we must critically rethink evaluation paradigms and actively embrace model heterogeneity as a core design principle.