arXiv cs.AI· Mingxuan Li, Huanzhi Mao, Ruitao Zheng, Jarod Stanbury James, Yunze Li, Natanya Anderson, Joseph E. Gonzalez, Alex Dimakis·· 11 小时前AI 评分48
SlackBench：在真实代码仓库协作项目上评测 AI 智能体的基准
SlackBench: Benchmarking Agents on Collaborative Projects Grounded in Real Code Repositories
AI 导读
SlackBench 是一个评测 AI 智能体在真实科研与工程协作工作流中表现的基准，基于真实学术项目的隐私保护重演环境，包含 140 条查询，覆盖异构来源检索、不可答问题弃答、目标导向摘要和长对话情境推断。在 33 个前沿模型与智能体框架上，最佳系统得分约 80%，多数智能体低于 66%。数据集已公开。
正文 · 原文
Abstract:AI Agents are becoming digital coworkers that can browse the web, write code, run experiments, and assist humans over extended horizons. Yet existing benchmarks fall short of real knowledge work: many rely on synthetic data or isolated tasks, missing the complex, multi-party dynamics of collaborative projects, where critical information is scattered across conversations, code, and evolving decisions. Even simple workplace questions may require locating relevant channels and direct messages, recovering context from long discussions, inspecting repository state, reconciling stale or conflicting evidence, and synthesizing across sources.
We introduce SlackBench, a benchmark for evaluating agents on realistic research and engineering workflows that require joint reasoning over workplace communication and evolving repository state. Grounded in privacy-preserving reenactments of real academic research projects, SlackBench provides reproducible environments with realistic multi-party dialogue, evolving decisions, and corresponding code changes grounded in GitHub repositories and Google Docs. It contains 140 queries spanning heterogeneous-source agentic search, abstention on unanswerable or false-premise questions, goal-directed summarization, and circumstance inference over long conversations. These tasks are not reducible to static corpus question answering: agents must determine which conversations matter, traverse threads, inspect repository state, and synthesize evidence across communication and code artifacts. Across 33 frontier models and agent harnesses, the best system scores around 80% and most agents score below 66%, highlighting substantial headroom for improvement. SlackBench thus provides an extensible framework for evaluating agents in collaborative work environments. The SlackBench dataset is available at this https URL .
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.05512 [cs.AI] |
| (or arXiv:2611.05512v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05512 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mingxuan Li [view email]
[v1]
Tue, 3 Nov 2026 19:10:50 UTC (8,048 KB)
来源：arXiv cs.AI · arxiv.org