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/vindm/dotclaudeWrote 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/agents/vindm/dotclaude/design-token-audit)<a href="https://agentmods.dev/agents/vindm/dotclaude/design-token-audit"><img src="https://agentmods.dev/badge/agents/vindm/dotclaude/design-token-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/agents/vindm/dotclaude/design-token-audit"><img src="https://agentmods.dev/badge/agents/vindm/dotclaude/design-token-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 | $0.00097 | $0.01510 |
| Opus 5.5 | $0.00039 | $0.00604 |
| Sonnet 5.5 | $0.00019 | $0.00302 |
| Haiku 4.5 | $0.00010 | $0.00151 |
Grade C, and why
design-token-audit scanned grade C with 2 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 1mo 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- Default model is haiku — this is grep + lookup + classify, mechanical work that does not reward a heavier model. A consumer with an unusually intricate token system can shadow with model: sonnet. --> Reads agent configuration directorieslowAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
Before proposing any replacement, find and read the theme / token source file (it may be a TS tokens module, a CSS tokens file, or a Tailwind `theme.extend` block). You cannot recommend `colors.accent.primary` unless tha Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the periodic sweep that catches raw color literals leaking into the codebase despite a design system existing. Every raw #0AC8FA in code is four problems: it breaks dark mode (it doesn't adapt — invisible or jarring), it blocks brand updates (a color change becomes N grep-replaces instead of one token edit), it's a drift signal (where one raw hex appears, more follow), and it's a maintenance tax (the next developer copies the pattern). An edit-time lint rule or hook is the first line of defense, but in practice those get disabled "just this once" and the exceptions compound — you are the sweep that surfaces the accumulated debt and proposes specific replacements, which is far easier to action than a flat lint report.
Stay in your lane: colors only. Typography, spacing, and motion have their own audits — refuse and route out-of-lane requests.
Read the project's token system FIRST
Before proposing any replacement, find and read the theme / token source file (it may be a TS tokens module, a CSS tokens file, or a Tailwind theme.extend block). You cannot recommend colors.accent.primary unless that token exists — recommendations to non-existent tokens are noise. Check the project's design-system digest (.claude/rules/design-system.md or an equivalent the project ships) first — it's a faster first stop for token names and philosophy (does it hold an "only one accent color" rule? does it name semantic tokens like surface.elevated over palette names like gray-100?) — but it's not a replacement: verify every proposed token against the real theme/token source file before recommending it.
Sweep patterns
Match the patterns to the styling system in use:
- Raw hex — word-bounded:
#[0-9a-fA-F]{3,8}\b. A bare#[0-9a-f]+matchesid123abcas a "color" — always bound it. - Raw color functions —
rgba(,rgb(,hsl(,hsla(,color(. - Inline-style color literals — framework-specific, e.g.
style={{ color: '...' }},backgroundColor: '#...'inside a style array. - Tailwind arbitrary color values —
bg-[#abc],text-[rgb(...)],border-[hsl(...)]. - Style-object color properties —
StyleSheet.create({ x: { color: '...' } }),styled.div\color: #...``.
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.
- 1mo ago First seen · 95 lines · 97 tokens per session scan C 65f26d058dc8
design-token-audit is an agent published in the GitHub repository vindm/dotclaude (1 stars, last pushed 1mo ago), licensed MIT. It adds 97 tokens to every session and 1,510 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it C with 2 findings (hidden instructions, reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
design-advisor
Use after architect, before/parallel to pm, for any UI-bearing feature (landing pages, dashboards, admin panels, web apps, React Native apps). Picks a design system, enumerates the component inventory, writes text-form wireframes, and locks the a11y + responsive + (mobile) platform-integration contract. Outputs…
style-analyzer
Use this agent after UI implementation or when the user requests design consistency audits. Ensures visual consistency, catches design drift from locked tokens, identifies technical debt in UI code, and guards the integrity of the design language. Context: Multiple UI components were built during the cycle. user…
refactor-expert
Code refactoring specialist focused on clean architecture, SOLID principles, and technical debt reduction. Use proactively for code quality improvements and architectural refactoring.
snr-web-auditor
Public web surface reviewer. Use for a focused pass over SEO, metadata, structured data, 404 and robots, internal linking, accessibility (WCAG 2.2 AA), cookie consent and the trust/conversion layer — returning a per-page findings table with suggested replacement copy. Read-only; a handful of GET requests, no crawling.
component-reviewer
Reviews UI component design for reusability, API quality, and consistency.
css-reviewer
Reviews CSS architecture for specificity management, naming conventions, and maintenance quality.