Dryui Build Experiment D logo

Dryui Build Experiment D

Community
rob-balfre
dryui-build-experiment-d

Experimental DryUI UI build workflow using branch routing, page-shell container queries, visual checks, and agentic variant selection. Use when testing whether the model can create, score, and choose among layout variants before implementation.

Overview

Publisherrob-balfre
Repositorydryui
Skill namedryui-build-experiment-d
Stars
9
Forks
0
Bundled files
9
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 9 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by rob-balfre on GitHub. Read the source before you install it.

Installation

Install the Dryui Build Experiment D AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/rob-balfre/dryui/tree/main/reports/skill-experiments/2026-05-11/variant-d-agentic-gated
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/rob-balfre/dryui.git /tmp/dryui
mkdir -p .claude/skills
cp -r /tmp/dryui/reports/skill-experiments/2026-05-11/variant-d-agentic-gated .claude/skills/dryui-build-experiment-d
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Dryui Build Experiment D in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Dryui Build Experiment D on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Dryui Build Experiment D is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

DryUI Build Experiment D: Agentic Gated Workflow

Target Router

Before implementation, identify the target:

  • Full page: route, screen, dashboard, admin page, settings page, docs page.
  • Section: hero, header, sidebar, table area, pricing block, footer.
  • Repeated component: card, row, list item, toolbar item, table cell.
  • Form/workflow: settings form, wizard, editor, onboarding, destructive action.
  • Polish pass: spacing, hierarchy, states, responsive behavior, visual refinement.

If the target is a dashboard or data surface, apply Dashboard Branch first, then Full Page Branch when it is a route or screen.

Target Brief Gate

Write this before code:

text
Target brief: type=<page|section|component|form|polish>, subtype=<dashboard|docs|marketing|settings|other>, user=<role>, primary_task=<task>, density=<compact|standard|spacious>, required_states=<states>.

Agentic Variant Gate

When the harness supports subagents and the user has asked for exploration, run three parallel variants before committing:

  1. Structure variant: optimize information architecture, regions, and responsive order.
  2. Density variant: optimize dashboard density, scan speed, and data hierarchy.
  3. State/resilience variant: optimize loading, empty, error, long labels, dense rows, and mobile collapse.

Each variant must return:

  • Mobile, tablet, and desktop layout plan.
  • data-layout / data-layout-area structure.
  • src/layout.css structure.
  • Risk list against DryUI lint and visual checks.

Then score each variant from 1-5 on:

  • Task fit.
  • DryUI contract.
  • Mobile-first/container-query quality.
  • Visual hierarchy.
  • State coverage.
  • Implementation risk.

Pick one winner or merge the strongest parts. Do not average weak variants into a muddled design.

If subagents are unavailable, create the three variants locally as short sketches, then score them before implementation.

Full Page Branch

For a full page:

  1. Create a page shell with a specific data-layout="<page-name>-shell".
  2. In src/layout.css, put container: <page-name> / inline-size on the shell.
  3. Put the responsive grid on an inner child, usually data-layout-area="page".
  4. Write @container <page-name> (...) rules that style descendants inside the shell; do not rely on a container query to style the container element itself.
  5. Start with mobile-first single-column layout.
  6. Add tablet and desktop shifts with @container <page-name> (min-width: ...).
  7. Use data-layout-area for stable regions and named grid areas.
  8. Keep src/layout.css structural only: display, grid, flex, container, tokenized spacing, alignment, and block-size constraints.
  9. Keep color, background, border, shadow, radius, typography, position, z-index, width, height, and inline-size out of src/layout.css.

Pattern:

svelte
<main data-layout="analytics-dashboard-shell">
	<section data-layout-area="page">
		<header data-layout-area="header">...</header>
		<section data-layout-area="filters">...</section>
		<section data-layout-area="metrics">...</section>
	</section>
</main>
css
[data-layout='analytics-dashboard-shell'] {
	container: analytics-dashboard / inline-size;
}

[data-layout='analytics-dashboard-shell'] > [data-layout-area='page'] {
	display: grid;
	gap: var(--dry-space-4);
	grid-template-areas: 'header' 'filters' 'metrics';
}

@container analytics-dashboard (min-width: 48rem) {
	[data-layout='analytics-dashboard-shell'] > [data-layout-area='page'] {
		grid-template-columns: repeat(2, minmax(0, 1fr));
	}
}

Dashboard Branch

For a dashboard or data surface:

  1. Pick the primary task: monitor, compare, triage, report, configure, or act.
  2. Pick density: compact for admin/data, standard for mixed dashboards, spacious only for executive summaries.
  3. Define regions before markup:
    • header
    • filter/action bar
    • summary metrics
    • primary visualization
    • data table or list
    • secondary detail/action area if needed
  4. Put filters/actions before the data they affect.
  5. Give the largest stable region to the primary chart, table, or work surface.
  6. Use Table or DataGrid for tabular data, and Chart or Sparkline for visualized data.
  7. Use cards only for KPIs, alerts, or distinct comparable units. Do not scaffold the whole page with cards.
  8. Include loading, empty, error, disabled, and dense-data states.
  9. Test long labels, many rows, narrow containers, and no-hover/coarse-pointer behavior.
  10. Make sure the primary action remains visible after mobile collapse.

Visual Check Gate

For visual work, verify screenshots at:

  • Mobile: 390px wide.
  • Tablet: 820px wide.
  • Desktop: 1440px wide.

Pass criteria:

  • No horizontal page overflow.
  • No clipped or overlapping text.
  • Primary task and action are visible in the first viewport.
  • Visual hierarchy survives mobile, tablet, and desktop.
  • Tablet is not just a stretched phone layout when a better two-column structure fits.
  • Desktop does not waste space with full-width KPI bands unless intentionally report-like.
  • Empty, loading, error, disabled, and dense-data states keep stable structure.
  • If adaptive theme is enabled, check light and dark; otherwise keep the app light-only.

Self-Review Gate

Before finishing, answer:

  • Did the correct target branch run?
  • Did the page shell own the named container?
  • Did @container rules style descendants inside the shell, not the shell itself?
  • Are page-level layout rules only in src/layout.css?
  • Are responsive shifts mobile-first with tablet and desktop checkpoints?
  • Does src/layout.css avoid visual styling and width/height/inline-size hacks?
  • Were mobile, tablet, and desktop screenshots checked?
  • Are required loading, empty, error, disabled, and dense-data states represented?
  • Would DryUI lint reject any markup, component usage, or CSS rule?

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Dryui Build Experiment D AI skill do?

Experimental DryUI UI build workflow using branch routing, page-shell container queries, visual checks, and agentic variant selection. Use when testing whether the model can create, score, and choose among layout variants before implementation.

Why use Dryui Build Experiment D on TypingMind?

Because you install it once and use it with any model. Dryui Build Experiment D is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Dryui Build Experiment D in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/rob-balfre/dryui/tree/main/reports/skill-experiments/2026-05-11/variant-d-agentic-gated. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Dryui Build Experiment D?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Dryui Build Experiment D?

As many as you like. As long as a model supports skills, you can use Dryui Build Experiment D with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Dryui Build Experiment D AI skill free?

Yes. It is published on GitHub by rob-balfre under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇