Impeccable Lite logo

Impeccable Lite

Community
ilindaniel
impeccable-lite

Design, build, redesign, critique, audit, and refine frontend interfaces with strong visual direction and production-ready UX. Use for websites, landing pages, dashboards, application UI, components, forms, onboarding, empty states, responsive behavior, accessibility, typography, color, layout, motion, UX copy, or when an interface feels generic, bland, cluttered, inconsistent, or unfinished. Do not use for backend-only or non-UI tasks.

Overview

Publisherilindaniel
Repositoryimpeccable-lite
Skill nameimpeccable-lite
Stars
27
Forks
3
Bundled files
Instructions only
LicenseApache-2.0
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by ilindaniel on GitHub. Read the source before you install it.

Installation

Install the Impeccable Lite 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/ilindaniel/impeccable-lite/tree/main
  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/ilindaniel/impeccable-lite.git \
  .claude/skills/impeccable-lite
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Impeccable Lite 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 Impeccable Lite 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 Impeccable Lite 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.

Impeccable Lite

Derived and modified from Impeccable by Paul Bakaus under Apache-2.0.

Create distinctive, coherent, production-ready interfaces without adding process machinery. Apply design judgment directly to the user's request and the existing product.

Principles

  • Let the brief win. Honor explicit aesthetics, constraints, content, brand decisions, and user priorities.
  • Preserve product truth. Never invent claims, capabilities, prices, testimonials, metrics, or configuration.
  • Distinguish refinement from redesign. Refinement preserves the current identity and behavior. Redesign preserves the product and function but may replace the visual language.
  • Treat existing code as evidence. Read the relevant route, component, tokens, styles, and representative neighboring UI before deciding.
  • Prefer one committed direction over a mixture of safe ideas.
  • Make design serve the surface. Marketing can earn expression; product UI should disappear into the task.
  • Reuse the project's components, tokens, icon set, and conventions before introducing anything new.
  • Keep verification bounded. Build fully, inspect once, fix findings together, confirm once, and stop.

Workflow

  1. Understand the request and inspect the target plus one representative source of visual truth.
  2. Classify the work as creation, redesign, refinement, critique, or technical audit.
  3. Identify the surface mode and the visitor's primary success condition.
  4. Resolve only decisions that materially change the result. Ask concise questions when audience, content, scope, or visual authority is genuinely unclear.
  5. Commit to a coherent direction before editing.
  6. Implement the complete requested scope, including meaningful states and responsive behavior.
  7. Verify the rendered result and relevant automated checks in no more than two inspection rounds.
  8. Report what changed, what was verified, and anything intentionally left out.

For critique or audit requests, inspect and report without editing unless the user also asks for changes.

Surface modes

Choose by surface, not by company category.

Persuade

Use for landing pages, campaigns, pricing, and marketing. Help the visitor understand, believe, and act. Give the first viewport a clear thesis, demonstrate the product instead of repeating claims, and make the primary action unmistakable.

Operate

Use for dashboards, tools, settings, editors, and authenticated application UI. Optimize for task completion, scanability, consistency, familiar affordances, and complete interaction states. Decoration must not compete with the work.

Read

Use for documentation, articles, guides, help, and changelogs. Optimize structure, wayfinding, typography, and comprehension. Keep prose readable and navigation predictable.

Experience

Use for portfolios, galleries, showcases, and immersive work. Let the artifact lead while the interface recedes. Use interaction and motion only where they strengthen exploration.

Establish direction

For an established surface, inherit its visual system unless the user explicitly requests a redesign. A missing design document does not make an existing product greenfield.

For a new surface, determine:

  • who is using it and in what situation
  • what they need to understand or accomplish
  • what content and functionality are real
  • what visual world fits the subject and audience
  • what should make the result recognizable rather than generic
  • what must remain unchanged

Define the direction in one compact statement covering the visual thesis, palette character, typography character, composition, and one memorable interaction or detail. Do not present a large menu of near-identical options when one well-supported direction will do.

Craft the interface

Hierarchy and layout

  • Make the primary action and information hierarchy obvious within seconds.
  • Group related elements tightly and separate distinct sections generously.
  • Use spacing rhythm deliberately, with more space above a heading than below it.
  • Prefer composition and alignment over wrapping everything in containers.
  • Use cards only when grouping benefits from a bounded surface. Never nest cards by reflex.
  • Make responsive changes structural: collapse navigation, reorder content, resize grids, and adapt tables instead of merely shrinking type.
  • Avoid horizontal overflow at every supported viewport.

Typography

  • Choose type that fits the product rather than defaulting to fashionable sameness.
  • Use a clear scale with intentional differences in size, weight, and line height.
  • Keep long-form body text near 65 to 75 characters per line.
  • Test the real copy. Handle long names, translated text, dynamic values, and user content without clipping.
  • Use monospace for code, data, or measurement, not as a costume for technical products.

Color and depth

  • Build a small semantic palette for backgrounds, surfaces, text, borders, actions, and states.
  • Maintain at least 4.5:1 contrast for normal text and 3:1 for large text and essential graphical controls.
  • Derive secondary text from its surrounding color context instead of placing generic gray on colored surfaces.
  • Use accent color to communicate priority or state, not to decorate every section.
  • Pick either a border or a shadow as the primary elevation cue. Avoid redundant outlines under diffuse shadows.
  • Use gradients, glass, glow, grain, and blur only when the chosen visual world requires them.

Components and interaction

  • Preserve semantic HTML and native platform behavior where possible.
  • Use the existing component library before creating replacements.
  • Provide visible hover, focus, active, selected, disabled, loading, empty, success, warning, and error states where relevant.
  • Keep keyboard navigation and screen-reader relationships intact.
  • Make touch targets at least 44 by 44 CSS pixels when practical.
  • Keep overlays usable outside clipping containers and within the viewport.
  • Prefer inline or progressive interactions over a modal unless interruption or protected focus is necessary.
  • Write controls as actions and errors as a problem plus a recovery path.

Motion

  • Use motion to explain state, continuity, hierarchy, or causality.
  • Keep content visible by default and honor prefers-reduced-motion.
  • Favor a small number of authored moments over identical entrance animations on every section.
  • Keep product UI transitions quick and avoid making frequent tasks wait for choreography.
  • Avoid bounce and elastic easing unless the product's established motion language explicitly calls for it.

Content and assets

  • Use the product's vocabulary and preserve factual copy unless changing it is in scope.
  • Prefer real or clearly labeled illustrative content over vague placeholders.
  • Use one consistent icon family. Do not substitute emoji or arbitrary Unicode symbols for interface icons.
  • Use meaningful imagery at the scale the composition needs. Do not fill missing content with decorative chrome.
  • Give every informative image useful alternative text and every decorative image empty alternative text.

Reject generic defaults

Unless the brief or existing system clearly earns them, avoid:

  • repetitive grids of equal cards with an icon, heading, and paragraph
  • nested cards and containers inside containers
  • generic hero layouts that could belong to any product
  • decorative metric cards, progress rings, or charts without information value
  • gradient text and purple-blue gradients as automatic emphasis
  • a rounded-square icon tile above every heading
  • an eyebrow label above every section title
  • oversized pills for ordinary containers or controls
  • decorative side borders on cards and callouts
  • glassmorphism, grid backgrounds, or noise added only to make a surface feel designed
  • display fonts in dense application controls
  • custom interactions that make standard tasks less familiar
  • invented testimonials, numbers, integrations, or product behavior

Do not merely soften a generic pattern. Replace it with a composition grounded in the content, task, and visual direction.

Production floor

Do not trade away:

  • semantic structure and accessible names
  • keyboard operation and visible focus
  • readable contrast in light and dark themes
  • reduced-motion support
  • responsive behavior across mobile and desktop
  • loading, empty, error, disabled, and success states where applicable
  • input validation and safe content rendering
  • internationalization and text expansion resilience
  • image sizing, lazy loading, and stable layout
  • cross-browser behavior for supported browsers
  • performance appropriate to the effect's value

Avoid adding a dependency for a minor visual effect. Prefer CSS, native browser capabilities, and installed project tools. Bound expensive animation and rendering work, clean up observers and listeners, and avoid blocking initial content on non-critical assets.

Verification

For frontend changes:

  1. Run the project's relevant lint, type, and test commands.
  2. Inspect desktop and mobile together in one batched round using available browser or screenshot tooling.
  3. Check hierarchy, overflow, contrast, controls, focus, states, content truth, and consistency with the chosen direction.
  4. Fix material findings in one batch.
  5. Confirm the fixes once and stop polishing.

For critique or audit work, report skipped or failed browser steps with concrete reasons.

Frequently asked questions

What does the Impeccable Lite AI skill do?

Design, build, redesign, critique, audit, and refine frontend interfaces with strong visual direction and production-ready UX. Use for websites, landing pages, dashboards, application UI, components, forms, onboarding, empty states, responsive behavior, accessibility, typography, color, layout, motion, UX copy, or when an interface feels generic, bland, cluttered, inconsistent, or unfinished. Do not use for backend-only or non-UI tasks.

Why use Impeccable Lite on TypingMind?

Because you install it once and use it with any model. Impeccable Lite 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 Impeccable Lite in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ilindaniel/impeccable-lite/tree/main. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Impeccable Lite?

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 Impeccable Lite?

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

Is the Impeccable Lite AI skill free?

Yes. It is published on GitHub by ilindaniel under the Apache-2.0 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 👇