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 arcasilesgroup/ai-engineering --skill ai-audit-designgit clone --depth 1 https://github.com/arcasilesgroup/ai-engineeringWrote 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/arcasilesgroup/ai-engineering/ai-audit-design)<a href="https://agentmods.dev/skills/arcasilesgroup/ai-engineering/ai-audit-design"><img src="https://agentmods.dev/badge/skills/arcasilesgroup/ai-engineering/ai-audit-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/arcasilesgroup/ai-engineering/ai-audit-design"><img src="https://agentmods.dev/badge/skills/arcasilesgroup/ai-engineering/ai-audit-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 | $0.00092 | $0.01311 |
| Opus 5.5 | $0.00037 | $0.00524 |
| Sonnet 5.5 | $0.00018 | $0.00262 |
| Haiku 4.5 | $0.00009 | $0.00131 |
Grade A, and why
ai-audit-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 19d 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 — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ai-audit-design
A stylesheet is a claim. A painted pixel is the evidence. Everything here measures the second, because the defects that survive review are exactly the ones no declaration predicts: contrast against a gradient with grain over it, a row whose two cards start reading 68px apart, a label printing over another label at 320px, a card grid that groups the wrong things.
Serve the built output and point the audit at it. A dev server with hot reload measures a page nobody visits.
node scripts/audit.mjs --base http://127.0.0.1:4399 \
--routes / /about /projects --widths 320 390 820 1440
--checks selects passes (geometry contrast collision proximity type interior a11y); --aa relaxes the contrast target from AAA to AA; --json out.json writes findings and text anchors. Run it before touching anything.
What each pass is asking
geometry — does the page fit, can a thumb hit it, does a heading follow the one above it. Sub-pixel borders on a pill radius, images shipping more resolution than their box can use, missing dimensions.
contrast — the ink is read from the element's own color; only the ground is sampled, because that is the part no token knows. Text over a photograph is where this earns its keep.
collision — glyph rects, not element boxes. Boxes overlap constantly and mean nothing.
proximity — for a set of like siblings, the gap between them against the largest gap inside one. Below 1.0 the space is grouping the wrong things and something else — usually a border — is arguing against it. Also inventories every stroke that is not a container edge.
type — one role, one answer. Two roles at the same size, face, weight and case whose leading differs by a hundredth is a typed value, not a decision.
interior — do containers in one row start reading on the same line, and does a component's inset scale with the type it holds.
a11y — structural and semantic checks: valid semantic markup, ARIA attribute validity and required attributes, heading / landmark / skip-link / lang / title structure, zoom disabled, deprecated elements, list-integrity (ul/ol/li, dt/dd/dl), focusable-aria-hidden mismatch, duplicate ids and accesskeys, form label coverage, and caption requirements. The a11y pass runs once per unique route across all widths; it is not width-dependent.
What ships with it
5 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.
- 19d ago Changed f51a5283fb47
- 22d ago First seen · 84 lines · 92 tokens per session scan A 5b88aa9926ee
ai-audit-design is a skill published in the GitHub repository arcasilesgroup/ai-engineering (61 stars, last pushed 2d ago), licensed Apache-2.0. It adds 92 tokens to every session and 1,311 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-09-24.
Other skills, from other repositories
review
WCAG 2.2 AA curated accessibility review — distinguishes automatically detectable, browser-assisted, and manual/human-judgment findings with evidence citations and SC mapping. Composes with design-assessment/design-improvement/frontend-design-review.
web-design-guidelines
Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".
ui-ux-design
Create production-grade frontend interfaces with strong UX and visual craft. Use when building web components, pages, dashboards, forms, landing pages, modals, or any UI. Use when user says 'build a form', 'create a dashboard', 'design a component', 'make a landing page', or asks for UI/UX work. Use even when the user…
frontend-design-review
Review and create distinctive, production-grade frontend interfaces with high design quality and design system compliance. Evaluates using three pillars: frictionless insight-to-action, quality craft, and trustworthy building. USE FOR: PR reviews, design reviews, accessibility audits, design system compliance checks…
visual-artifact-qa
Visual output that passes static checks can still fail to render.
frontend-design
Use this skill when building UI components, designing frontend architecture, implementing styling systems, optimizing performance, ensuring accessibility, or creating production-grade interfaces. Trigger on keywords: frontend, component, UI, styling, React, Next.js, Angular, Tailwind, CSS, responsive, accessibility…