RESEARCHInvestigateNEXT 12 MONTHS
CoSA: The Conflict of Skills Attack in Real-World LLM Agent Ecosystems
arXiv cs.CL — Computation and Language
Factual evidence
What the source reports
Researchers identify Conflict of Skills (CoS), a vulnerability where composition of safe LLM agent tools causes policy violations.
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 27 October 2026
- Collected by OneBench
- 28 Oct 2026, 02:01 UK
Stored source excerpt
arXiv:2610.35082v1 Announce Type: new Abstract: Agent Skills allow LLM agents to combine reusable capabilities, but checking each Skill in isolation cannot rule out policy violations…
Short excerpt from the collected text, not the full source. Use the source link to read it in context.
The factual summary is a OneBench synthesis, not a quotation or independent verification. Collection time is not publication time. Open the source for its full context; related reporting can share the same underlying announcement.
OneBench interpretation
Institutional assessment
So what
Multi-tool agent workflows can bypass security boundaries even if individual tools pass isolated safety checks.
Do what
Review agentic workflow security architecture to ensure access policies are evaluated across multi-tool chains rather than in isolation.