SkillAgentSearch skills...

prototype-spec-html

An AI agent skill that turns requirement descriptions into shareable HTML prototypes with element-level annotations

Install / Use

npx skills add Paritck/prototype-spec-html

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

71/100

Category

Design

Supported Platforms

Universal

Tags

Our assessment of prototype-spec-html

prototype-spec-html scores 71/100 on our quality scale, 245th of 269 Design skills we index.

Its SKILL.md is 6.8 KB long, well organised into 15 sections with 3 code examples: a thorough specification that gives an agent plenty to work with.

It has 2 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
29/30
Structure
18/20
Description
12/15
Adoption
2/20
Freshness
11/15

Maintenance, license and trust

  • The repository was last updated about 3 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
  • Our last check on 2026-10-19 found the source still online.
  • It is released under the Apache-2.0 license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 92/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-09-24. Automated pattern scan on 2026-09-24. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

prototype-spec-html compared with similar skills

All 4 of these similar skills score higher than prototype-spec-html; compare them before choosing.

SkillScoreStarsUpdatedFormat
prototype-spec-html (this skill)by Paritck7123mo agoSKILL.md
algorithmic-artby anthropics100177.9k45d agoSKILL.md
pptxby anthropics100177.9k45d agoSKILL.md
designby nextlevelbuilder100143.4k6d agoSKILL.md
ui-ux-pro-maxby nextlevelbuilder100143.4k6d agoSKILL.md

Frequently asked questions

How do I install prototype-spec-html?
Run npx skills add Paritck/prototype-spec-html. The install tabs above show the steps for each supported agent.
Which AI agents does prototype-spec-html work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is prototype-spec-html safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is Apache-2.0-licensed and scores 92/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is prototype-spec-html still maintained?
The repository was last updated about 3 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.

prototype-spec-html

An open-source AI agent skill that turns a product manager's natural-language requirement description into a single shareable, offline-viewable HTML file containing a three-column prototype workspace with element-level requirement annotations.

What It Does

Give the AI a requirement description like:

做一个订单管理后台,含订单列表页和订单详情页。列表页有搜索框、状态筛选、表格(订单号/客户/金额/状态/时间/操作)和分页;详情页展示订单信息、商品明细和物流跟踪。

The skill generates one .html file with a three-column workspace:

  • Fixed header — project / system / requirement name + generation time.
  • Left sidebar — clickable page list with hash routing (#page-order-list), each page showing a Web or Mobile badge.
  • Center canvas — AI-drawn high-fidelity wireframe prototype with lightweight interactions (hover, collapse, tabs, dialogs). Each page is designed for either Web size (full-width admin layout) or Mobile size (centered phone-frame shell with notch, status bar, and bottom tab bar).
  • Right sidebar — element-level requirement annotations linked bidirectionally to the prototype.

Web-Size Page

Web-size page

Web-size pages fill the canvas width with a desktop admin interface: internal sidebar + topbar + content area, just like a real management system.

Mobile-Size Page

Mobile-size page

Mobile-size pages center a 390px phone-frame shell inside the canvas, complete with notch, status bar, stacked card lists, and a bottom tab bar. No <table> — mobile pages use card lists exclusively.

Element Picking Mode

Picking mode

When adding or editing an annotation, the user clicks "选择元素" (Pick Element) to enter picking mode: the canvas dims, and as the user moves the mouse over the prototype, the element under the cursor is outlined in real time. Clicking an element selects it and auto-fills the data-spec field — no manual typing required. Elements without a data-spec get one assigned automatically.

Bidirectional Linkage

Linkage highlight

Hovering a prototype element highlights both sides simultaneously — the element itself and its corresponding annotation (yellow background + scroll into view). Hovering an annotation likewise highlights both the annotation and its prototype element. Clicking an annotation highlights both sides and scrolls the prototype element into view. Linkage is suppressed during picking mode to avoid interference.

Key Features

  • Single file, zero dependencies — all CSS, JS, and data are inlined. No CDN, no frameworks, no build step. Double-click to open in any browser.
  • Per-page size design — each page is designed for Web or Mobile at design time (no runtime viewport switcher). A single HTML can mix both.
  • Element-level annotations — every annotation is linked to a specific prototype element via data-spec / id matching, not just page-level.
  • Runtime annotation editing — add, edit, and delete annotations directly in the right sidebar using Markdown. An inlined lightweight Markdown renderer (headers, tables, bold, italic, inline code, lists) renders the content to HTML. Changes persist to localStorage and survive page refresh.
  • Element picking — pick a prototype element by mouse (hover to preview, click to select) instead of typing data-spec values. User-picked data-spec attributes also persist across refreshes.
  • Cross-tool compatible — works with any agent tool that supports the standard SKILL.md format.

Install

This skill follows the standard SKILL.md format and works with any agent tool that supports it.

Option A: via skills CLI

npx skills add https://github.com/Paritck/prototype-spec-html

Option B: manual copy

Copy the prototype-spec-html/ folder (containing SKILL.md and references/) into your agent's skills directory:

| Agent tool | Project-level path | Global path | |---|---|---| | Claude Code | .claude/skills/ | ~/.claude/skills/ | | Codex CLI | .agents/skills/ | ~/.agents/skills/ | | Cursor | .cursor/skills/ | ~/.cursor/skills/ | | GitHub Copilot | .github/skills/ | ~/.copilot/skills/ | | Windsurf | .windsurf/skills/ | ~/.codeium/windsurf/ | | Goose | (per Goose docs) | — | | Zed | (per Zed docs) | — |

Cursor and GitHub Copilot also scan .claude/skills/ for compatibility, so a single copy there covers three tools.

Usage

After installation, just describe your requirement to the AI in any supported tool:

用 prototype-spec-html 生成一个后台管理系统原型:
- 项目名称:客户管理平台
- 页面1(Web):客户列表(搜索、表格、分页)
- 页面2(Web):客户详情(基本信息、订单记录、跟进记录)
- 页面3(Mobile):App首页(快捷入口、最近订单)

The AI will load the skill, follow the 5-step workflow, and deliver a single .html file.

Repository Structure

prototype-spec-html/
├── SKILL.md                              # skill entry: frontmatter + workflow + constraints
├── references/
│   ├── html-template-spec.md             # three-column layout + CSS + JS + element picking
│   └── prototype-prompt-guide.md         # AI drawing rules + Web/Mobile + annotation points
├── demo/
│   ├── order-management-prototype.html   # demo output (open in browser to see it live)
│   └── screenshots/                      # screenshots for README
├── README.md                             # English readme (this file)
├── README_CN.md                          # Chinese readme
└── LICENSE                               # Apache 2.0

How Element-Level Linkage Works

  • Prototype elements carry data-spec="annot-001".
  • Right-sidebar annotation blocks carry id="annot-001".
  • Vanilla JS (inlined) listens for mouseenter and click:
    • Hover a prototype element → both the element and its annotation highlight (yellow) and the annotation scrolls into view.
    • Hover an annotation → both the annotation and its prototype element highlight (yellow).
    • Click an annotation → both sides highlight and the prototype element scrolls into view.
  • Switching pages clears all highlights.
  • During picking mode, linkage is suppressed so the user can select elements without interference.
  • Newly added annotations (via the "选择元素" button) automatically participate in linkage.
  • User-picked data-spec attributes persist to localStorage and are restored after refresh.

Demo

Open demo/order-management-prototype.html directly in a browser to see a live example with 4 pages:

  1. 订单列表 (Web) — admin table with filter bar, order table, pagination
  2. 新建订单 (Web) — form with basic info, product selection, action buttons
  3. App首页 (Mobile) — mobile home with banner, quick actions, order cards
  4. 订单详情 (Mobile) — mobile order detail with status steps, order info, action buttons

License

Apache 2.0 — see LICENSE for details.

Related Skills

View on GitHub
GitHub Stars2
CategoryDesign
Updated3mo ago
Forks0

Languages

HTML

Trust signals

92/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 low