Skillware：让智能体不再依赖提示词的模块化可审计能力执行标准
Don't Prompt Your Agents. Equip Them: Skillware as a Modular Standard for Sovereign, Auditable, and Deterministic Capability Execution for Logical Systems
AI 导读
arXiv 论文提出 Skillware，一个将智能体“想做什么”与“如何执行”分离的开放架构框架和能力注册表，用模块化、版本化的软件能力包取代松散提示词。在 23 个前沿与开源权重模型、368 个基准条件下，配备 Skillware 的 3B-8B 参数本地小模型在复杂多步工作流中稳定超过未配备的前沿模型，token 用量最多减少 84.8%，推理成本最多降低 97.3%，并完全消除参数错误。
正文 · 原文
Abstract:Current autonomous logical systems rely on a fragile assumption: that language models should dynamically write code, guess API payloads, and memorize operational constraints through prompt engineering. In practice, this burns massive context windows, inflates API bills, and produces brittle agents that fail unpredictably in production.
In this work, we present Skillware, an open architectural framework and capability registry that formally separates what an agent wants to do from how that task is executed. Instead of prompting models with loose text instructions or markdown files baptized as 'skills', Skillware equips them with modular, versioned software capability bundles. Much like skill hotkeys in video games that execute deterministically under fixed world rules, each bundle combines a typed JSON input contract, deterministic Python code, task directives, bundled datasets, and automated unit tests. In-process adapters translate these capabilities into native tool declarations for any model, from frontier cloud APIs to local open-weight runtimes with zero network latency.
Evaluating 23 frontier and open-weight models across 368 benchmark conditions, we show that small local models (3B to 8B parameters) equipped with Skillware consistently outperform unequipped frontier models on complex multi-step workflows. They cut token usage by up to 84.8%, slash inference costs by up to 97.3%, and eliminate parameter errors entirely. True agent reliability does not come from larger models or longer prompts, but from equipping logical systems with verified, deterministic software.
Finally, we articulate the trajectory of capability bundles across an invariant continuum of information processing systems: from digital logical systems today, to IoT, drones, and embodied robotics tomorrow, establishing a foundation for auditable, real-time capability execution.
| Comments: | 51 pages, 26 figures. Code repository and benchmarks available at this https URL |
| Subjects: | Artificial Intelligence (cs.AI); Software Engineering (cs.SE) |
| ACM classes: | I.2.11; D.2.11; I.2.7 |
| Cite as: | arXiv:2611.07652 [cs.AI] |
| (or arXiv:2611.07652v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07652 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Vladimiros Peilivanidis [view email]
[v1]
Thu, 5 Nov 2026 08:38:51 UTC (5,397 KB)
来源：arXiv cs.AI · arxiv.org