Overview
AI agents are rapidly evolving from experimental prototypes to always-on services that autonomously plan, invoke external tools, collaborate, and continuously interact with their environment. This shift challenges traditional operating system abstractions—processes, threads, files, sockets, and resource controllers—which were never designed for dynamic, semantically rich, adaptive agent workloads. To support AI agents at scale, operating systems themselves must become agentic, adapting their abstractions and resource management policies to the semantic behaviors of agents.
The AgenticOS workshop seeks original position papers and experience reports that explore OS-level mechanisms for AI-agent workloads. Our goal is to define the primitives, isolation models, scheduling techniques, and observability mechanisms necessary to build operating systems explicitly tailored to agent-based systems.
Workshop Schedule
March 23, 2026 — Afternoon Session
| Time | Session | Details |
|---|---|---|
| 1:30pm – 1:35pm | Opening | Workshop Opening Remarks |
| 1:35pm – 2:00pm | Keynote | Keynote Talk (25 min) |
| 2:00pm – 2:15pm | Invited Talk | Invited Talk #1 (15 min) |
| 2:15pm – 2:30pm | Research Paper |
AgentCgroup: Understanding and Controlling OS Resources of AI Agents
Yusheng Zheng, Jiakun Fan, Quanzhi Fu, Yiwei Yang, Wei Zhang and Andi Quinn
|
| 2:30pm – 2:45pm | Research Paper |
Rethinking OS Interfaces for LLM Agents
Yuan Wang, Mingyu Li and Haibo Chen
|
| 2:45pm – 3:00pm | Research Paper |
Skills are the new Apps — Now It's Time for Skill OS
Le Chen, Zichang Wang, Wenxin Zheng, Erhu Feng, Dong Du, Yubin Xia and Haibo Chen
|
| 3:00pm – 3:15pm | Research Paper |
Toward LLM-Driven Rule Generation for Enforcement Systems: An Exploratory Study on WAF
Quanzhi Fu and Dan Williams
|
| 3:15pm – 3:45pm | Break | Poster & Coffee Break (30 min) |
| 3:45pm – 4:00pm | Invited Talk | Invited Talk #2 (15 min) |
| 4:00pm – 4:15pm | Research Paper |
Fork, Explore, Commit: OS Primitives for Agentic Exploration
Cong Wang and Yusheng Zheng
|
| 4:15pm – 4:30pm | Research Paper |
It is Time to Virtualize Foundation Models with a Self-evolving Operating System Layer
Suparna Bhattacharya, Tarun Kumar, Cong Xu, Satish Kumar Mopur, Jiahao Li, Ashish Mishra, Aalap Tripathy, Annmary Justine Koomthanam, Martin Foltin and Ian Foster
|
| 4:30pm – 4:45pm | Research Paper |
Fuyun: Bridging the Semantic Gap in Serverless Resource Provisioning via LLM Agents
Qingwen Li, Kai Lv, Mingxuan Yang, Zhengyu Lei, Cunchi Lv, Xiao Shi and Xiaofang Zhao
|
| 4:45pm – 4:50pm | Vision Paper |
Mobile-MCP: Implementing the Model Context Protocol via Android's Inter-Application Communication Mechanisms
Xiheng Li, Mengting He, Chengcheng Wan and Linhai Song
|
| 4:50pm – 4:55pm | Vision Paper |
pMVX: Policy-Level Multi-Version Execution for Agentic OS Kernel Self-Tuning
Sujot Singh, Eddie Federmeyer, Kenan Alghythee and Xiaoguang Wang
|
| 4:55pm – 5:00pm | Vision Paper |
Execute-Only Agents: Architectural Defense Against Prompt Injection for AI Agents
Rahul Tiwari and Dan Williams
|
| 5:00pm – 5:05pm | Vision Paper |
Towards Agentic Performance Management
Jingyuan Chen, Gongqi Huang and Amit Levy
|
| 5:05pm – 5:10pm | Vision Paper |
Grimlock: Guarding High-Agency Systems with eBPF and Attested Channels
Gan Fang, Sheng Mao, Mark Wu, Wenhui Zhang, Biao Gao, David Levitsky, Shawna Murphy Butterworth and Rob Cameron
|
| 5:10pm – 6:00pm | Panel | Panel Discussion (50 min) |
| 6:30pm | Social | Social Dinner |
Topics of Interest
Topics of interest include (but are not limited to):
- New OS abstractions for agent execution (process/container/multikernel enhancements)
- Dynamic sandboxing and lightweight runtimes for securely executing agent-generated code and tasks
- Semantics-aware resource management and scheduling for dynamic, multi-agent workloads
- Long-lived state abstractions for managing agent context, prompts, and episodic memory
- eBPF-driven extensions for real-time observability, adaptation, and constraint enforcement
- Compiler–OS co-design for adaptive and agent-aware JIT execution strategies
- GPU and accelerator virtualization for large-scale deployment of agent workloads
- Security and isolation mechanisms for agent-invoked tools, code, and data flows
- Agents managing systems: kernel tuning, anomaly detection, resource orchestration, failure recovery, and dynamic policy updates
Submission Guidelines
We solicit two types of submissions:
Vision Papers
1–2 pages (excluding references).
Suitable for early-stage ideas, position statements, ongoing projects, demos, and insights from production systems. We strongly welcome contributions from industry practitioners and the open source community to share real-world experiences and challenges.
Research Papers
Up to 6 pages (excluding references).
Suitable for more complete concepts, research results and experience reports.
All submissions must follow the ACM double-column conference format. The review process is single-blind, with each submission receiving at least two reviews.
Frequently Asked Questions
Can I submit work that is already on arXiv?
Yes. Posting on arXiv or other preprint servers does not disqualify a submission.
Can I submit work-in-progress or concurrent submissions?
Yes. We welcome early-stage ideas and ongoing work. Submissions under review at other venues are acceptable.
Will this affect future publication at other venues?
No. This workshop has no formal proceedings. Accepted papers will only appear on the workshop website, which does not preclude future publication at conferences or journals.