摘要：我们探讨了由本地部署大语言模型驱动的 AI Agent，是否能够在贴近行业实际的工具调用环境中，可靠地自动化由专家定义的硬件设计工作流。在这些环境中，工程师通过专用工具执行重复且存在依赖顺序的操作——例如创建组件、添加端口和连接线路。组件规格与命名约定通常受保密限制，因而无法使用托管式专有 API，这也推动了本地部署模型的使用。为研究这一场景，我们构建了一个模型上下文协议（MCP）服务器，用于复现嵌入式系统开发中某专有硬件设计工具的状态和依赖逻辑；同时构建了一项基准测试，涵盖单操作编辑、多步骤依赖链、无效请求、拼写错误提示以及多服务器工具上下文。我们评估了七个开源模型，并比较了包括系统提示、工具描述详细程度、上下文范围，以及单 Agent 与多 Agent 架构在内的不同流水线选择。结果表明，强大的模型可在所测工作流中实现接近完整的预期调用覆盖率，但其可靠性同时强烈依赖任务结构和 Agent 配置。完整的工具描述能够持续减少失败；少样本提示会使部分模型出现严重的不作为；累积上下文会损害受限模型的表现；多 Agent 拆分则能够帮助能力较弱的工作 Agent 或支持较长会话，但代价是增加额外调用。这些发现为在有状态的硬件设计环境中部署本地 LLM Agent 提供了实用指导。

Abstract:We ask whether AI agents powered by locally deployed large language models can reliably automate expert-defined hardware design workflows in an industry-realistic tool-calling setting. In these environments, engineers issue repetitive, dependency-ordered operations---such as creating components, adding ports, and wiring connections---through specialised tools. Confidentiality constraints on component specifications and naming conventions often preclude hosted proprietary APIs, motivating the use of locally deployed models. To study this setting, we build a Model Context Protocol (MCP) server that reproduces the state and dependency logic of a proprietary hardware design tool used in embedded system development and construct a benchmark covering single-operation edits, multi-step dependency chains, invalid requests, misspelled prompts, and multi-server tool contexts. We evaluate seven open-source models comparing pipeline choices including system prompts, tool-description detail, context scope, and single-agent versus multi-agent architectures. Results show that strong models can achieve near-complete expected-call coverage on the benchmarked workflows, but reliability depends strongly on both task structure and agent configuration. Comprehensive tool descriptions consistently reduce failures, few-shot prompting can cause severe inaction for some models, cumulative context harms constrained models, and multi-agent decomposition helps weak workers or long sessions at the cost of additional calls. These findings provide practical guidance for deploying local LLM agents in stateful hardware design environments.