Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward

Renjun Xu, Yang Yan

Keywords: Agent skills, large language models, tool use, computer-use agents, Model Context Protocol, skill libraries, agentic AI, progressive disclosure, security

TL;DR: Agent skills are powerful, but community-contributed skills have vulnerabilities, motivating our proposed four-tier Skill Trust and Lifecycle Governance Framework that maps skill provenance through verification gates to graduated runtime permissions

Abstract:

The transition from monolithic language models to modular, skill-equipped agents marks a defining shift in how large language models (LLMs) are deployed in practice.
Rather than encoding all procedural knowledge within model weights, agent skills---composable packages of instructions, code, and resources that agents load on demand---enable dynamic capability extension without retraining. It is formalized in a paradigm of progressive disclosure, portable skill definitions, and integration with the Model Context Protocol (MCP).
This survey provides a comprehensive treatment of the agent skills landscape, as it has rapidly evolved during the last few months.
We organize the field along four axes: (i) architectural foundations, examining the SKILL.md specification, progressive context loading, and the complementary roles of skills and MCP; (ii) skill acquisition, covering reinforcement learning with skill libraries (SAGE), autonomous skill discovery (SEAgent), and compositional skill synthesis; (iii) deployment at scale, emphasizing computer-use agents (CUAs) while situating them alongside software-engineering, web, tool-use, and embodied-agent benchmarks; and (iv) security, where recent empirical analyses report that 26.1\% of community-contributed skills contain vulnerabilities, motivating our proposed Skill Trust and Lifecycle Governance Framework---a four-tier, gate-based permission model that maps skill provenance to graduated deployment capabilities.
We identify seven open challenges---from cross-platform skill portability to capability-based permission models---and propose a research agenda for realizing trustworthy, self-improving skill ecosystems.
Unlike prior surveys that broadly cover LLM agents or tool use, this work focuses specifically on the emerging skill abstraction layer and its implications for the next generation of agentic systems.
An accompanying collection of agent skills resources is available at \url{https://github.com/scienceaix/agentskills}.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 9