---
title: "UniClawBench：面向真实世界任务的主动代理通用基准"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.08768v1"
permalink: "https://feed.astralor.com/items/itm_06a3358d54809db66745e760"
published: "2026-07-09T17:59:32.000Z"
score: 79
channel: "ai"
tags: ["HKU-MMLab","UniClawBench","主动代理","基准测试","多模态大语言模型","大语言模型"]
---

# UniClawBench：面向真实世界任务的主动代理通用基准

## AI 点评

> 这篇论文提出的UniClawBench正是我们急切需要的——它不再假设完美环境，而是把代理丢进真实的Docker容器，用细粒度检查点考验实际能力。对做Agent框架和模型评估的读者来说，这是必读的参考基准。

## 速览

UniClawBench是首个能力驱动的基准，用于评估在动态真实环境中操作的主动代理。它围绕五项基础模型能力（技能使用、探索、长上下文推理、多模态理解、跨平台协调）构建了400项双语真实任务。与传统静态基准不同，UniClawBench在实时Docker容器中评估代理，并采用分层检查点和闭环评估策略，模拟多轮人类反馈。研究通过多种代理框架对比，揭示了基础模型能力与框架设计如何共同影响真实世界性能。

## 关键点

- 提出五项核心能力维度，系统评估主动代理在真实场景中的表现。
- 采用实时Docker容器和逐步骤完成检查点，替代传统静态预录答案。
- 设计闭环评估策略，模拟真实多轮用户反馈且不泄露评分标准。
- 通过多模型与多框架对比，解耦基础模型能力与框架设计的影响。

## 全文

The rapid development of large language models and multimodal large language models has accelerated the emergence of proactive agents capable of operating everyday tools and assisting users in real-world environments. However, existing benchmarks struggle to evaluate such agents effectively, as they often rely on sandboxed environments and single-turn evaluation paradigms. Moreover, their scenario-based task taxonomies mix multiple model capabilities within the same task category, making it difficult to identify the root causes of agent failures. To address these limitations, we introduce UniClawBench, the first capability-driven benchmark designed to evaluate proactive agents in dynamic, real-world settings. UniClawBench is built around five foundational model capabilities: Skill Usage, Exploration, Long-Context Reasoning, Multimodal Understanding, and Cross-Platform Coordination. Based on these capabilities, we design 400 bilingual real-world tasks. Unlike previous benchmarks that rely on static, pre-recorded answers, our benchmark evaluates agents in live Docker containers using fine-grained, step-by-step completion checkpoints. Furthermore, we design a closed-loop evaluation strategy comprising an executor agent, a hidden supervisor agent, and a user agent to simulate realistic multi-turn human feedback without leaking grading criteria. To disentangle base model capabilities from framework-level design choices, we evaluate state-of-the-art models under multiple agent frameworks. Through comprehensive comparisons across both models and frameworks, we show how base model capabilities and agent framework designs jointly shape performance in real-world environments. To facilitate future research, we make our benchmark and code publicly available at https://github.com/HKU-MMLab/UniClawBench.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_06a3358d54809db66745e760