Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add sanky369/vibe-building-skills --skill original-designgit clone --depth 1 https://github.com/sanky369/vibe-building-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/sanky369/vibe-building-skills/original-design)<a href="https://agentmods.dev/skills/sanky369/vibe-building-skills/original-design"><img src="https://agentmods.dev/badge/skills/sanky369/vibe-building-skills/original-design/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sanky369/vibe-building-skills/original-design"><img src="https://agentmods.dev/badge/skills/sanky369/vibe-building-skills/original-design.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00153 | $0.03383 |
| Opus 5 | $0.00077 | $0.01691 |
| Sonnet 5 | $0.00031 | $0.00677 |
| Haiku 4.5 | $0.00015 | $0.00338 |
Grade A, and why
original-design scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 221 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Original Design
Run an interactive, question-driven session that turns a product or subject into a complete design language — a coherent visual world from which every screen, frame, icon, illustration, and animation is drawn as an "artifact" — and deliver it as a filled-in Design Language Brief. It works for any visual deliverable: an infographic, a video storyboard, a website, or a mobile app UI/UX. The governing principle: originality is not about being unique — it is about being internally consistent around one core idea. Don't ask "how do I make this look unique?"; ask "what is the visual manifestation of the core idea?" The biggest shift this skill enforces: don't design screens — design a world. Everything else is an element of that world.
When to use / when not to
- Use when the user needs an original visual identity or art direction, or an existing design reads generic/AI-default and needs one unifying idea.
- If the user needs a practical brand style guide (palette, fonts, mood for asset generation) rather than an original world →
skills/creative/creative-strategist; this skill's brief can feed that one. - If the design language already exists and the user just needs assets → go straight to the production skill for the medium (see Phase 9).
- If the problem is UI structure/usability rather than visual identity →
skills/frontend-design/frontend-orchestrator.
How to run this skill (read first)
This is a facilitated conversation, not a form dump. Follow these rules:
- One phase at a time. Ask the focused question(s) for the current phase, wait for the answer, reflect it back in one or two sentences, then move on. Never paste all seven phases at once.
- Always offer a recommendation. For every question, propose 2–4 concrete options (with a "(recommended)" first) so the user can react instead of inventing from scratch. They can always answer freely.
- Use structured questions when the host supports them. In Claude Code, prefer the
AskUserQuestiontool for option-based phases. In other agents, ask in plain text with a short labeled option list. - Gather live visual inspiration from the web at Phase 1.5 and Phase 6 (see references/inspiration-browsing.md). Pick the browsing method that matches the host agent.
- Capture decisions as you go into a running Design Language Brief (template at assets/design-language-brief-template.md).
- Gate on the Originality Test before producing final output. Don't skip it.
- Hand off to a production skill for the chosen medium (see references/medium-playbooks.md).
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 221 lines · 153 tokens per session scan A 319c07401833
original-design is a skill published in the GitHub repository sanky369/vibe-building-skills (30 stars, last pushed 2mo ago), licensed MIT. It adds 153 tokens to every session and 3,383 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
mermaid
Create, validate, and repair Mermaid.js diagrams. Use when generating flowcharts, sequence, class, ER, state, or Gantt diagrams, or any visualization.
lfe-visual-check
Inspector sub-skill. Renders the changed UI surface, reasons over the result, and writes a text findings record with a human-action instruction and sign-off token — so a human visual sign-off can close a UI change. Artifact-free (saves no image files). Auto-arms on a visual-file touch; written by the Inspector at…
accessibility-compliance-accessibility-audit
You are an accessibility expert specializing in WCAG compliance, inclusive design, and assistive technology compatibility. Conduct audits, identify barriers, and provide remediation guidance.
avalonia-layout-zafiro
Guidelines for modern Avalonia UI layout using Zafiro.Avalonia, emphasizing shared styles, generic components, and avoiding XAML redundancy.
brand-guidelines
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
brand-guidelines
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.