TMLR 论文：仅靠终端与文件系统的终端智能体足以完成企业自动化

View PDF HTML (experimental)

Abstract:There has been growing interest in building agents that can interact with digital platforms to execute meaningful enterprise tasks autonomously. Among the approaches explored are tool-augmented agents built on abstractions such as Model Context Protocol (MCP) and web agents that operate through graphical interfaces. Yet, it remains unclear whether such complex agentic systems are necessary given their cost and operational overhead. We argue that a coding agent equipped only with a terminal and a filesystem can solve many enterprise tasks more effectively by interacting directly with platform APIs. We evaluate this hypothesis across diverse real-world systems and show that these low-level terminal agents match or outperform more complex agent architectures at a fraction of the cost. Our findings suggest that simple, flexible programmatic interfaces combined with strong foundation models should be the backbone of enterprise automation.

Comments:	Published in Transactions on Machine Learning Research (TMLR), 10/2026. 56 pages, 6 figures, 23 tables. Code: this https URL. Data: this https URL. Reviews: this https URL
Subjects:	Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
Cite as:	arXiv:2604.00073 [cs.SE]
	(or arXiv:2604.00073v4 [cs.SE] for this version)
	https://doi.org/10.48550/arXiv.2604.00073 arXiv-issued DOI via DataCite
Journal reference:	Transactions on Machine Learning Research, 2026

Submission history

From: Patrice Bechard [view email]
[v1] Tue, 31 Mar 2026 14:14:47 UTC (5,130 KB)
[v2] Fri, 3 Apr 2026 15:48:53 UTC (5,130 KB)
[v3] Wed, 5 Aug 2026 16:41:57 UTC (4,294 KB)
[v4] Wed, 14 Oct 2026 14:23:23 UTC (2,793 KB)