By WishingCat
Automate Ansys HFSS/AEDT antenna and RF simulation workflows via PyAEDT, including environment setup, geometry modeling, boundary/port assignment, solver configuration, frequency sweeps, S-parameter extraction, far-field plots, and parametric optimization.
A Claude Code / Codex CLI skill for driving Ansys HFSS antenna & RF simulations through PyAEDT.
让 AI agent 用 PyAEDT 驱动 Ansys HFSS 跑天线/RF 仿真。装上之后,只要你跟 agent 说"用 HFSS 仿真这个天线"、"扫频 2-6 GHz 提 S 参数"、"画方向图",它就会按这个 skill 给的标准工作流来做。
pyaedt-hfss-antenna/
├── SKILL.md # 主入口:何时触发 + 9 步标准工作流
├── references/
│ ├── environment-setup.md # 版本对照、安装、gRPC/COM、license
│ ├── antenna-workflow.md # 主线 9 步的 API 详解
│ └── common-pitfalls.md # 高频报错速查
├── templates/
│ ├── sparam_extraction.py # S 参数 + 谐振点 + -10 dB 带宽
│ ├── far_field_pattern.py # 远场方向图 + 增益 + 辐射效率
│ └── parametric_sweep.py # 变量参扫(native + Python loop)
├── scripts/
│ └── check_env.py # 探测本机 AEDT/Python/pyaedt
└── .claude-plugin/plugin.json # Claude Code 插件清单
设计原则:进度式披露(progressive disclosure)。SKILL.md 给主线,具体步骤需要时再去 references / templates 里查,不会一次性把所有内容塞进 agent 上下文。
| 你的 Ansys AEDT/HFSS | 推荐 PyAEDT | Python | 接口 |
|---|---|---|---|
| 2026 R1 | >= 0.24 | 3.10–3.13 | gRPC(默认) |
| 2025 R2 | >= 0.18 | 3.10–3.12 | gRPC |
| 2025 R1 | 0.14–0.17 | 3.9–3.12 | gRPC |
| 2024 R2 | 0.10–0.13 | 3.9–3.12 | gRPC |
| 2024 R1 | <= 0.9.x | 3.8–3.11 | gRPC |
| 2023 R1 / R2 | 0.6–0.9 | 3.8–3.10 | gRPC |
| 2022 R2 | 0.5–0.7 | 3.7–3.9 | gRPC 可用 |
| 2022 R1 | 0.5–0.7 | 3.7–3.9 | 仅 COM |
AEDT 在 macOS 上无原生支持,Mac 用户需远程到 Windows/Linux 服务器,本地 PyAEDT 走 gRPC 连接。详见 references/environment-setup.md。
把这个仓库直接克隆到 Claude Code 的全局 skills 目录:
# macOS / Linux
git clone [email protected]:WishingCat/pyaedt-hfss-antenna.git \
~/.claude/skills/pyaedt-hfss-antenna
# Windows (PowerShell)
git clone [email protected]:WishingCat/pyaedt-hfss-antenna.git `
"$env:USERPROFILE\.claude\skills\pyaedt-hfss-antenna"
之后在任何 Claude Code 会话里,只要你的请求与 HFSS / 天线仿真 / pyaedt 相关,这个 skill 就会自动出现在可用列表里被触发。
也可以装成 Claude Code 插件(支持 marketplace 风格):
# 仓库根带 .claude-plugin/plugin.json,可被 Claude Code plugin 系统识别
# 通过 /plugin install 或将仓库加为 marketplace 后选择安装
不想全局生效,只想在某个项目目录里启用:
cd /path/to/your/project
git clone [email protected]:WishingCat/pyaedt-hfss-antenna.git \
.claude/skills/pyaedt-hfss-antenna
git clone [email protected]:WishingCat/pyaedt-hfss-antenna.git \
~/.codex/skills/pyaedt-hfss-antenna
Codex 与 Claude Code 都遵循
SKILL.md的 frontmatter 约定(name + description),仓库结构通用。
如果你不想往全局目录里装,可以把仓库克隆到任意位置,然后在会话里告诉 agent:
Please read the skill at /path/to/pyaedt-hfss-antenna/SKILL.md
and follow its workflow for any HFSS / antenna simulation task.
在装完之后,跑一下环境检测:
python ~/.claude/skills/pyaedt-hfss-antenna/scripts/check_env.py
期望输出:
============================================================
PyAEDT environment check
============================================================
[OS] Windows ...
[Python] 3.11.x
[pyaedt] 0.18.x
--- AEDT installations ---
v252 (2025 R2)
path : C:\Program Files\AnsysEM\v252\Win64
pyaedt 推荐版本:>=0.18
如果 [pyaedt] NOT INSTALLED,装一下:
pip install "pyaedt[all]"
装好后随便起一段对话:
"我有一个 2.45 GHz 的微带贴片天线工程在 D:\sim\patch.aedt,帮我扫频 2-3 GHz 提取 S11、谐振频率和 -10 dB 带宽。"
agent 会:
SKILL.md 找到对应模板 → templates/sparam_extraction.pyPROJECT_PATH / DESIGN_NAME / FREQ_* 参数references/antenna-workflow.md 了解 API 细节references/common-pitfalls.md更多触发短语:
MaximumPasses=3, num_of_freq_points=21)走通流水线,再加密。release_desktop(),否则 license 会泄漏。模板里都已经用 try/finally 包好。ansysedt.exe -grpcsrv <port> 启动 server。详见 references/common-pitfalls.md。
cd ~/.claude/skills/pyaedt-hfss-antenna
git pull
发现模板在你机器的 PyAEDT 版本上签名不对、或踩到 common-pitfalls.md 没覆盖的坑,欢迎开 issue 或 PR。
特别欢迎:
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub wishingcat/pyaedt-hfss-antennaLLM-driven RF control and analysis — "Claude Code for RF". Survey spectrum, build and run receivers, simulate scenes, and run closed-loop TX/RX experiments through natural language, leaving a reproducible RF project behind.
Foundational Model-Based Design capabilities for the Simulink product family
AI-powered hardware development platform — design, verify, synthesize, and deploy working RTL with natural language. 18 agents, 25 skills, 8 IP blocks.
97-agent RTL design pipeline with 97 skills for 6-Phase hardware design automation (Research → Architecture → μArch → RTL → Verify → Design Note)
18 specialized Claude Code skills for computational design in AEC covering parametric modeling, generative design, computational geometry, algorithmic patterns, structural computation, environmental simulation, facade engineering, digital fabrication, BIM scripting, interoperability, optimization methods, data-driven design, mesh processing, design automation, and machine learning for AEC. References 50+ pioneers and frameworks including Patrik Schumacher, Greg Lynn, Neri Oxman, Achim Menges, and standards from buildingSMART, Ladybug Tools, ASHRAE, Eurocode, and more. Includes 7 Python calculators.
Control KLayout from AI agents via MCP — geometry, display, image overlay, visual capture, nanodevice flake detection, GDS alignment, and routing skills for chip/mask layout design