Proteus Skills
A skill that teaches an agent to build and edit Proteus circuits, load microcontroller firmware and verify simulation results.
- 104 ★ on the repository
- Last commit
Extended access
Why this level
- The agent runs Python scripts and drives an installed Proteus.
- On first run the package is installed from the author's GitHub repository.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/proteus-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent runs Python scripts and drives an installed Proteus.
- On first run the package is installed from the author's GitHub repository.
This is third-party code. Review the repository files before installing.
What it does
The `proteus-skills` skill drives Proteus through the `proteus-automatic-api` Python library. The agent creates and opens `.pdsprj` projects, changes components, properties and wiring, then saves and reopens the project to verify it. It then exports an SDF netlist, loads ELF or HEX firmware into a microcontroller, runs a timed simulation and reads GPIO logs. Buttons and switches bind to two-state controls: the agent presses, releases and checks the circuit's response. Waveforms are read from existing graphs and probes and exported to CSV. The `scripts/validate_evidence.py` script separates structural checks from real runtime runs.
Who it is for. Electrical-engineering students and engineers who simulate circuits in Proteus on Windows and want an agent to handle the routine.
Good fit when
- Build an STM32 project where a button lights an LED and releasing it turns it off.
- Check an existing circuit's connections against an SDF netlist.
- Run a HEX firmware and read GPIO pin states.
- Capture a waveform from an existing probe and export it to CSV.
Not a fit when
- You only need theory or a circuit illustration: the authors exclude such requests.
- You run macOS or Linux: the skill targets Windows with Proteus installed.
- You expect the agent to click through missing steps in the Proteus window: GUI automation is forbidden, and the agent reports an API gap instead.
Example request
Use $proteus-skills: build an STM32 project where the LED is on while a button is pressed and show the simulation result.Limitations
Windows, an installed Proteus and component models are required. The workflow was verified on Python 3.12 and Proteus 8.16 SP3. Timed simulation and interactive controls depend on the Proteus DLL version, and other builds were not tested. On first run the agent installs the `proteus-automatic-api` package from the author's GitHub repository into `.venv-proteus`. Proteus itself, models and third-party firmware are not part of the repository. The author calls the library usable but unfinished.
How to disable. Delete the `proteus-skills` folder from the skills directory (`~/.codex/skills/proteus-skills/` by default) and the `.venv-proteus` virtual environment from the working folder.
FAQ
Do I need to install proteus-automatic-api in advance?
No. The skill checks Python and the package itself and, if missing, installs it from the author's repository into `.venv-proteus` before continuing.
What should I prepare before starting?
A `.pdsprj` project or template, the Proteus program path and an output folder. For a microcontroller add the firmware and target pins. A new circuit needs a template with component definitions, for example `Rescap.pdsprj`.
How do I tell a real simulation from a structural check?
The `validate_evidence.py` script checks the evidence file. The `--require-runtime` flag demands a confirmed run and rejects `runtime_verified` claims that carry runtime errors.
Which Python and Proteus versions were verified?
Python 3.12 and Proteus 8.16 SP3 (build 8.16.36097). Python 3.10 or newer is required, and timed simulation depends on the Proteus DLL version.
Which ready examples does the repository include?
Three STM32 button and LED scenarios: hold, click-to-toggle and two independent buttons. They come with C firmware sources, HEX files and build scripts.
Discussion
No questions or reviews yet.
Related
SkillsA skill that builds an EasyEDA Pro eprj3 project from a circuit description: schematic, board placement and automated gates.
A skill that walks an agent from a hardware idea to a PCB, verifies routing and stops before the order is paid.
24 skills for microcontroller firmware: build, flash, GDB debugging, UART, Modbus, CAN, logic analyzers and static analysis
A skill, CLI and MCP server for EasyEDA Pro: the agent reads datasheets, draws schematics and boards and checks them with DRC
A Claude Code skill for designing PCBs, CAD and firmware as code that writes fab files only after every gate passes
A Verilog-2001 design skill that writes RTL from a brief and explains and repairs existing modules with a check report
README in short
The repository holds the skill instructions and examples, while the library lives in a separate repository. Examples include three STM32 button and LED scenarios: hold, toggle and two independent buttons. The README gives Codex instructions, and the install path for Claude Code is the same.
README badge
Original SKILL.md
--- name: proteus-skills description: 使用 proteus-automatic-api Python 库创建、编辑和验证真实 Proteus/ISIS 工程，加载 MCU 固件、操作按钮/开关并读取仿真结果。用于用户要求 Proteus Skills、操作 .pdsprj、Proteus 自动化或在 Proteus 中测试按钮驱动电路的任务；普通电路讲解和概念配图不需要此技能。 license: MIT --- # Proteus Skills 通过 `proteus_automatic_api` 交付可继续编辑的 `.pdsprj` 及可核对的原生结果。技能名称是 **Proteus Skills**，调用名是 `$proteus-skills`。独立代码仓库与 Python 分发包均名为 **`proteus-automatic-api`**，Python 导入名为 `proteus_automatic_api`。 技能与库独立分发。本技能不包含库源码、wheel、Proteus 软件、官方模板或第三方器件/固件。仓库中的 [原创 STM32 演示固件](examples/stm32) 单独列明用途；它们不替代本机 Proteus 资源。使用技能不要求克隆库仓库、将两者放在相邻目录或从技能目录运行任务。 所有 Proteus 操作必须通过 `proteus_automatic_api` 的公开 API 完成，包括选件、放置、连线、属性配置、仿真和结果读取。不得使用 Computer Use、OCR、屏幕坐标、鼠标键盘自动化或其他 GUI 操作工具补做步骤，也不得直接调用库的私有 UI 助手替代缺失的公开 API。`Session` 等公开接口内部驱动真实 Proteus 属于库工作流。 库未覆盖的操作必须明确报告能力缺口，不能自动回退到 GUI 后仍称其为本技能完成的操作。录制工具只可被动采集库真实执行的过程，不得驱动应用或修改工程。用户要求从零录制时，应从空白工程开始，记录实际选件、逐个放置、逐段连线、配置和验证；不能把完成态截图拼接称为完整绘制过程。公开 API 无法实时显示某个构建步骤时，明确说明限制，不伪造过程。 文件编辑保存后可用新的 `Session` 展示该步骤。原始录制和事件日志保留会话关闭、启动及等待过程，以及 `Library.search/get` 的实际查询、结果和操作时间；不能伪装成图形选件操作。发布 GIF 或视频默认只使用原生界面截图和原生配置对话框，不擅自添加外框、标题、API/结果说明或其他多余叠字。可剪去没有匹配窗口的等待段及其时长，保留全部真实窗口帧和实际操作顺序，不展示等待卡；独立调试工具窗不叠放到原理图上。剪辑范围、加速倍率及额外停帧写在媒体之外的说明中，原始录制和日志保持完整。 ## 环境与版本 把“用户只有 Proteus，尚未安装 Python 库”作为正常首次使用场景。环境准备是必须执行的工作流，不是给用户阅读的前置条件。缺库时必须实际运行安装，验证成功后在同一任务中继续操作 Proteus；不能只给安装命令、等待用户手动安装或在安装成功后结束。 1. 检查 Windows、可用的 Python 3.10+ 和用户实际安装的 Proteus。优先使用任务已有虚拟环境；没有 Python 时协助从官方来源安装，不能把安装 Python 库说成安装了 Proteus 软件。 2. 在选定解释器中检查 `proteus_automatic_api` 的导入、版本、安装来源和本次所需接口。已有可用构建就继续；缺库时实际运行随技能提供的 [ensure_library.py](scripts/ensure_library.py)，不要仅展示命令；缺接口或用户要求更新时加 `--update`，详见 [首次安装与更新](references/distribution.md#首次安装与更新)。使用技能完成任务已包含为任务准备这个依赖的意图，不再询问是否要安装已声明的库；仍遵守宿主的执行权限。 3. 从固定的官方仓库 **https://github.com/kudoumakoto6523-design/Proteus_automatic_package** 查询默认分支最新提交，核对该提交的包名和公开接口，再安装其提交 SHA 对应的源码 ZIP。GitHub 是默认来源，不等待 PyPI、Release 或 wheel；不搜索同名仓库，不猜改名后的 URL，也不要求安装 Git。 4. 首次安装默认写入用户任务目录的 `.venv-proteus`，后续 pip、验证和任务脚本都使用同一解释器。安装失败不能进入电路操作；脚本成功后读取返回 JSON 中的 `python` 绝对路径，所有后续脚本用该解释器运行；检查本次所需 API 后立即继续用户的电路任务。不要在运行中的 Proteus 会话里换库，也不在每次 API 调用前检查更新。 在用户任务目录运行下面的命令；`<技能绝对路径>` 由 Agent 根据当前读取的本技能目录填写，不让用户寻找。示例解释器 `py -3.12` 换成已检测到的可用 Python。脚本只准备 Python 库，不启动或操作 Proteus。 ```powershell py -3.12 "<技能绝对路径>/scripts/ensure_library.py" ``` 返回 `action=installed` 或 `action=reused` 且进程退出码为 0 后，使用返回的 `python` 继续本次工作流。只有实际安装失败或确实缺少必要资源，才报告具体阻断；不能将“尚未安装库”本身当作阻断。 当前操作示例已在 **Python 3.12 / proteus-automatic-api 0.2.0 / Proteus 8.16 SP3（8.16.36097）** 验证。GitHub 新提交仍需按其文档核对接口和 Proteus DLL 兼容性；不能把相同版本号视为相同构建。README 演示所需的 `label_offsets` 及新建工程固件隔离语义须单独确认，历史演示构建证据见 [分发说明](references/distribution.md#验证范围)。 | 资源 | 0.2.0 内置默认值（使用前验证） | 配置入口 | |---|---|---| | Proteus | `D:\Proteus\BIN\PDS.EXE` | `Session(project, executable=...)`，要求 PDS.EXE | | 器件目录 | `C:\ProgramData\program\LIBRARY` | `Library(directory=...)`，仅查询 | | 新建模板 | `C:\ProgramData\program\SAMPLES\Graph Based Simulation\Rescap.pdsprj` | `Circuit(template_project=...)` | 按任务检查所需路径；库不会自动探测安装。`import_device(..., library=...)` 的 `library` 是库名选择器，导入仍使用默认目录。`Library(directory=...)` 不会改变导入器；可以从有对应定义的 donor 工程导入。空模板初始化也可能回退到默认 Rescap 样例。不要把这些局部路径参数描述为完整可移植配置。