Beginner Friendly Workbench Builder

Problem

Many personal tools are over-designed when treated as full management systems: expense tracking, habit check-ins, to-do lists, exam prep, fitness logging, kids’ learning, or creator workflow. Users often do not know how to specify fields, modules, colors, or prompts. Generic HTML generators also tend to produce pages that depend on external resources, feel poor on mobile, or lose data after refresh or device changes. This skill narrows the task to a shippable vertical use case: a zero-baseline user can describe a scenario and receive a working, portable page that can be deployed to the home screen.

How It Works

The skill routes keywords into nine built-in scenario presets, such as ledger workbench, fitness workbench, exam prep workbench, and personal IP workbench. It does not ask users to define fields, charts, or visual details first. Instead, it loads the expert specification from the matched preset, asks no more than two simple multiple-choice questions, and aims to start generation within two rounds.

The generated output follows explicit engineering constraints:

• A single HTML file with inline CSS and JavaScript, avoiding CDNs, external fonts, and chart libraries.
• Data stored in localStorage with wb_ prefixed keys, plus JSON export, import/restore, and a confirmation dialog before clearing; ledger-style outputs also support CSV export.
• Charts and icons rendered with inline SVG, avoiding Canvas libraries and emoji icons.
• Mobile-first layout for narrow screens, supporting deployment to a phone home screen.

The skill also includes assets/workbench-starter.html as a baseline template and references/verify_static.js as a static checker for hard rules such as zero external links, inline SVG, mobile adaptation, and wb_ key prefixes.

Boundaries

It is intended for personal use, not account systems, admin backends, browser push notifications, or large-file and video storage. The first version usually stays within three to four modules, so complex workflows should be iterated incrementally. If the need is highly unique and does not match a preset, the skill falls back to a guided workflow, but the goal remains to produce a minimal working version before adding more fields or visual details.