Token导航 LogoToken导航

Agent Skill · 研究检索

sensecraft-hmi-gen(sensecraft 人机界面)

sensecraft-hmi-gen 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

sensecraft-hmi-gen 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 12,173。

总安装 10,502周安装 429GitHub Stars 公开资料未说明下载 3,398核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:sensecraft-hmi-gen(sensecraft 人机界面)
来源仓库:https://github.com/killingjacky/sensecraft-hmi-gen
安装命令:
openclaw skills install sensecraft-hmi-gen
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install sensecraft-hmi-gen

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

为 SenseCraft HMI 电子墨水屏生成优化网页内容。AI 驱动布局选择兼顾阅读体验与低功耗显示。适合仪表盘、信息板与低功耗终端界面设计。sensecraft-hmi-gen 属于研究检索类 Skill,可作为该场景下的辅助能力补充。输出为轻量级 HTML/CSS,兼容常见浏览器渲染。

深色模式与低刷新率特性延长屏幕使用寿命。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install sensecraft-hmi-gen;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

sensecraft-hmi-gen 怎么安装?

在终端执行 openclaw skills install sensecraft-hmi-gen。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

sensecraft-hmi-gen 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

sensecraft-hmi-gen 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

sensecraft-hmi-gen 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

sensecraft-hmi-gen 用的人多吗?

公开数据是累计安装 10,502 次,近一周安装 429 次,下载 3,398 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

sensecraft-hmi-gen 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
sensecraft-hmi-gen

sensecraft 人机界面 · 本页

10,502—接触敏感数据MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
sensecraft-hmi-gen
description
Generate beautiful web content for SenseCraft HMI e-ink displays. AI-powered layout selection, e-ink optimization. Creates artistic, minimalist pages optimized for e-ink screens (800x480 to 1600x1200).

SenseCraft HMI Web Content Generator

Generate stunning, e-ink optimized web content with AI-powered design.

Philosophy

E-ink screens are digital art pieces, not web pages.

Inspired by Kindle screensavers and minimalist posters. Every page should be beautiful enough to frame.


AI Workflow

When the user requests to generate e-ink screen content, execute according to the following workflow:

Step 1: Study Reference Files and Keep Them in Mind

Read design reference files (mandatory):

Design Philosophy - {baseDir}/references/design-patterns.md

10 Classic Layout References - {baseDir}/references/layouts.md, but do not limit yourself to these layouts. Specifically, when the user provides layout ideas, follow the user's instructions.

Step 2: Guide the User to Answer Necessary Configuration Information

If the user is conversing with you via a webchat channel, call the wizard.js script. This script will serve a web page at http://localhost:39527/ and pop up this visual configuration interface in the user's browser. The script will automatically exit once the user completes the configuration. The script will create a {baseDir}/data/.wizard-config.json file to save the configuration information.

If the user is using a remote conversation channel like Discord, use dialogue to ask the user for the following information:

  • Screen size and color support
  • The user's ideas about the layout

These configuration details will be saved in the {baseDir}/data/.wizard-config.json file. Subsequent workflows should follow these configurations, such as the size of the generated web page, color support, etc.

Step 3: Prepare Content

Text Content:

  • AI generates content based on user requests (quotes, poems, weather info, etc.)
  • Or use content provided by the user

Image Content (if the user's request explicitly mentions the need for images):

Ask the user for the image source:

  • User provides image URL
  • User provides local image
  • If openclaw is configured with a text-to-image large model, call it to generate an image

Image storage path: {baseDir}/data/public/images/

Step 4: Generate HTML

Generate an HTML file based on the user's requirements and store it in the {baseDir}/data/ directory. The HTML file can contain Javascript, and the image paths referenced in the HTML should be ./public/images/.

Please embed CSS directly into the HTML file. If you need to reference a CSS framework like Bootstrap, use the CDN link for the CSS file.

Please embed Javascript scripts directly into the HTML file. If you need to reference a JS framework like jQuery, use the CDN link for the JS file.

Primary rules for the web page:

  • The pixel count of the entire web page's visible area must match the screen size exactly; there should be no scrolling controls.
  • The colors used in the web page should not exceed the color support capabilities of the screen. This includes controlling the background color, text color, image color, etc., using CSS.

Design philosophy and layout references:

  • {baseDir}/references/design-patterns.md
  • {baseDir}/references/layouts.md

Step 5: Start server.js

We use pm2 to manage the http server.

# Start service
pm2 start {baseDir}/data/server.js --name sensecraft-hmi

# Stop service
pm2 stop sensecraft-hmi

# View logs
pm2 logs sensecraft-hmi

# Access address
http://localhost:19527/?token=XXXXX

After the server starts, automatically open a browser to visit http://localhost:19527/?token=XXXXX so the user can see the generated web page. Also, prompt the user in the chat interface that they can open the http://localhost:19527/?token=XXXXX link to view the web page effect in real-time.

Step 6: Reverse Proxy

Guide the user to use a reverse proxy tool to forward http://localhost:19527 to the public network, and then use the html widget in the SenseCraft HMI platform to display the web page.

Reverse proxy tools include: frp, ngrok, cloudflare tunnel, etc. Encourage users to use HTTP forwarding rather than TCP forwarding because HTTP is more secure.

Due to openclaw skill security review reasons, this skill does not provide the binary of reverse proxy tools. Please ask the user to search, download, and use them on their own.


Tool List

ToolDescriptionUsage
init_project.jsProject initializationnode {baseDir}/scripts/init_project.js
wizard.jsConfiguration wizard servernode {baseDir}/scripts/wizard.js

10 Classic Layouts

See references/layouts.md for detailed specs.

LayoutIDUse Case
Minimalist CenterminimalCenterQuotes, maxims, single words
Info CardinfoCardPoetry, elegant text
Dashboard GriddashboardGridWeather, stocks, data
Comic FullscreencomicFullscreenComics, artwork
Calendar GridcalendarGridCalendars, schedules
News ListnewsListNews, RSS, to-do lists
Split LayoutsplitLayoutSidebar + main content
Big NumberbigNumberClocks, countdowns
Image & TextimageTextImages + captions
Newspaper FrontnewspaperFrontNews summaries

Scheduled Updates

If the user's request includes scheduled tasks, there are 3 approaches:

  1. Use OpenClaw cron to update content on a schedule, suitable when each update's content varies significantly
openclaw cron add --name "daily-summary" --schedule "0 9 * * *" \
  --task "Generate a work summary based on Notion"
  1. Embed Javascript scripts within the HTML file to dynamically update content, suitable for pulling data from APIs
  1. Write background services in a Node.js Express backend using templating technologies or frameworks like React to push data from backend to frontend, suitable for particularly complex content updates

Files Structure

sensecraft-hmi-gen/
├── SKILL.md                    # This file (current file)
├── data/                       # Source files for AI-generated web pages
│   ├── server.js               # Express server
│   ├── index.html              # Currently displayed content
│   ├── .token                  # Access token
│   ├── public/images/          # Image resources (optional)
│   ├── public/css/             # CSS resources (optional)
│   └── public/js/              # JS resources (optional)
├── scripts/
│   ├── init_project.js         # Project initialization
│   ├── wizard.js               # Configuration wizard server (Frontend + Backend)
└── references/
    ├── layouts.md              # Detailed layout specifications
    ├── design-patterns.md      # Design patterns and aesthetics
    └── screen-specs.md         # Hardware specifications

Tips

  • Let AI be creative - Don't force templates, let design flow from content
  • Read design references - design-patterns.md and layouts.md are your inspiration
  • Go bigger than you think - fonts that look huge on screen look perfect on e-ink
  • Test with actual device - e-ink looks different than LCD
  • Embrace whitespace - it's not wasted space, it's elegance

Troubleshooting

IssueSolution
Text too smallMinimum 32px, headings 60px+
Blurry imagesApply color dithering to images to make them better suited for e-ink displays
Crowded layoutIncrease padding to 40-60px
Service inaccessibleCheck token, run pm2 restart sensecraft-hmi

适合场景

01

文本生成图片

02

图片风格化

03

产品图和创意图

04

需要 FLUX 模型时

能力概览

能力 1

调用 FLUX 图像模型

能力 2

支持文本生图和图像改写

能力 3

覆盖 LoRA 或风格适配

能力 4

适合创意视觉生成

平台分布

OpenClaw

73.74%
按下载量换算2,506

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

未展示

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills