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.
git clone --depth 1 https://github.com/w2ur/claude-code-setupWrote 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/commands/w2ur/claude-code-setup/design-audit)<a href="https://agentmods.dev/commands/w2ur/claude-code-setup/design-audit"><img src="https://agentmods.dev/badge/commands/w2ur/claude-code-setup/design-audit/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/commands/w2ur/claude-code-setup/design-audit"><img src="https://agentmods.dev/badge/commands/w2ur/claude-code-setup/design-audit.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 | $0.00030 | $0.01288 |
| Opus 5.5 | $0.00012 | $0.00515 |
| Sonnet 5.5 | $0.00006 | $0.00258 |
| Haiku 4.5 | $0.00003 | $0.00129 |
Grade A, and why
design-audit 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 today.
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 — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit the live design of the portfolio. Audit only: nothing is fixed here. This command orchestrates scripts and a skill that already exist; it holds no site list, no viewport and no count. Each script is the source of truth for its own numbers.
Definitions used below: <date> is today as YYYY-MM-DD. <host> is the URL host. <slug> is <host> when the URL path is / or empty; otherwise <host>-<path>, the path lowercased with every run of characters outside a-z0-9 turned into one - and leading or trailing - trimmed (https://elenchus.my-bias-app.example.com/fr/ → elenchus.my-bias-app.example.com-fr), so two rows on one host never share a directory. <dir> is ~/.claude/state/design-audit/<date>/<slug>/.
1. Targets
$ARGUMENTS empty or all:
~/.claude/scripts/design-sites.sh
stdout is <repo>\t<url> per site (- in the repo column means no repo context). stderr is gap\t<repo> for each deployed repo with no homepage. Exit 0 complete, 1 gaps (the list is still printed), 2 unknown.
- Exit
2: stop. Say the site list is unknown and quote stderr. Never read it as "no sites". - Exit
1: continue, and list every gap at the top of the summary as a coverage hole.
$ARGUMENTS is a URL (starts with http): that single site, repo -, unless design-sites.sh prints that host with a repo.
$ARGUMENTS is a repo name: run design-sites.sh and keep that repo's row. No row means unknown for that repo; say so and stop.
2. Evidence phase, sequential
One Chrome at a time. Run this loop yourself, in the foreground, before any agent exists:
mkdir -p <dir> && rm -f <dir>/report.md
~/.claude/scripts/design-detect.sh <url> <dir> 2><dir>/detect.stderr
The rm clears a report left by an earlier run of the same day, so step 4 can only find a report this run wrote. detect.stderr keeps the script's explanation; the summary quotes it. Record each site's exit code. Exit 0 captured and clean, 1 captured with hits, 2 unknown (capture failed, blank or error page, detector crash). Do not parallelize this phase.
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.
- today First seen · 68 lines · 30 tokens per session scan A 4e6cfbdf39cd
design-audit is a command published in the GitHub repository w2ur/claude-code-setup (2 stars, last pushed yesterday), licensed MIT. It adds 30 tokens to every session and 1,288 once invoked, about $0.0001 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 2026-10-10.
Other commands, from other repositories
ui-synth
Master orchestration command that synthesizes ALL LibreUIUX plugins - archetypal-alchemy, design-mastery, accessibility, security, performance, testing - into one cohesive, production-ready UI/UX implementation. The ultimate synthesis.
ui-critique
Get comprehensive design feedback and improvement suggestions for UI components.
archetypal-ui
Generate a UI component from a Jungian archetype paired with a Major Arcana card.
style-guide
Generate a style guide with design tokens, type scale, spacing, components, and Tailwind or CSS config.
ui-modern
Generate a modern UI component with best practices and design system compliance.
design-audit
Audit a UI against eight design dimensions and return scored findings with fixes.