Terminal Agents Suffice for Enterprise Automation
Abstract
Simple terminal-based coding agents using programmatic interfaces and foundation models can effectively perform enterprise tasks comparable to or better than complex tool-augmented agents.
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. Our findings suggest that simple programmatic interfaces, combined with strong foundation models, are sufficient for practical enterprise automation.
Community
Terminal-based coding agents using direct API access can match or outperform GUI and tool-augmented agents for enterprise automation, suggesting strong foundation models with simple programmatic interfaces are often sufficient.
Hi @patricebechard , very cool paper!
Any reason not to use native computer use tooling like https://developers.openai.com/api/docs/guides/tools-computer-use, or the computer use tool built into the OpenAI Agents SDK? I assume it would work better than the Playwright MCP server (they might also be more expensive).
Interesting breakdown of this paper on arXivLens: https://arxivlens.com/PaperView/Details/terminal-agents-suffice-for-enterprise-automation-3452-111f1726
Covers the executive summary, detailed methodology, and practical applications.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Terminal Is All You Need: Design Properties for Human-AI Agent Collaboration (2026)
- Empirical Comparison of Agent Communication Protocols for Task Orchestration (2026)
- Building Effective AI Coding Agents for the Terminal: Scaffolding, Harness, Context Engineering, and Lessons Learned (2026)
- Beyond Context Sharing: A Unified Agent Communication Protocol (ACP) for Secure, Federated, and Autonomous Agent-to-Agent (A2A) Orchestration (2026)
- AgentRob: From Virtual Forum Agents to Hijacked Physical Robots (2026)
- AgentOS: From Application Silos to a Natural Language-Driven Data Ecosystem (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
terminal agents are underrated for enterprise stuff, good to see a paper making the case. the argument that you dont need more complex setups for most automation tasks is actually kind of compelling. detailed breakdown here https://arxivexplained.com/papers/terminal-agents-suffice-for-enterprise-automation
Terminal-based coding agents using direct API access can match or outperform GUI and tool-augmented agents for enterprise automation, suggesting strong foundation models with simple programmatic interfaces are often sufficient.
This is fascinating, will we see a ServiceNow CLI/TUI then?
Thanks for publishing!
Get this paper in your agent:
hf papers read 2604.00073 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper