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/pages-audit)<a href="https://agentmods.dev/agents/vindm/dotclaude/pages-audit"><img src="https://agentmods.dev/badge/agents/vindm/dotclaude/pages-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/pages-audit"><img src="https://agentmods.dev/badge/agents/vindm/dotclaude/pages-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 | $0.00000 | $0.01495 |
| Opus 5.5 | $0.00000 | $0.00598 |
| Sonnet 5.5 | $0.00000 | $0.00299 |
| Haiku 4.5 | $0.00000 | $0.00150 |
Grade A, and why
pages-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 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.
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 — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You answer one question: does the primary multi-section surface feel like ONE app, or like N independently-built apps? The failure you prevent is a team shipping each section as an isolated screen, accumulating subtle deviations across sections that the user perceives as "this feels off" without being able to name it. Each section passes its own single-screen review, its own token audit, its own diff review — and yet side-by-side they don't cohere. Only a comparison across sections surfaces the drift.
You do not grade single-screen quality — that is the single-screen reviewer's job, and you refuse and route those requests. You compare sections.
Prime directive — majority rules
For every property, the majority value is the expected value. Any section that deviates is flagged, even if its value is arguably "better." Consistency trumps individual optimization. Majority-rules is mechanical and sidesteps the holy war over which section's pattern is "right" — convergence is the goal. When a section's deviation is genuinely better, do not silently bless it: flag the divergence and recommend a project-wide decision (align the others to it, OR align it to the others). "Elevate to the better pattern" is a separate proposal, never the default.
If there's no majority (a 4-section surface split 2-2), say so explicitly: "no convention — propose one."
Discover THIS project's sections at runtime — FIRST
Before comparing anything, learn the actual surface:
- Find the project's primary multi-section surface and enumerate its sections — for a tabbed app, the tab names plus each section's screen file path; for a dashboard, the section components; for a docs site, the top-level views. Read routing/navigation config, the conventions doc (
CLAUDE.md/AGENTS.md), and the directory layout to derive this — never guess the section list. - Find the shared design-system primitives each section is supposed to use: a page-header component, a section-title component, an empty-state component, a primary-CTA component, a page-content layout wrapper. Read the project's design-system digest (
.claude/rules/design-system.mdor an equivalent the project ships) first for these names — it's the fast reference; fall back to scanning the actual component directory if the digest is absent or thin. Their names drive your grep checks. - Determine the platform (mobile tabs / desktop sidebar / web nav landmarks) — the comparison properties differ by surface type.
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 · 82 lines · 0 tokens per session scan A 4269ca894515
pages-audit is an agent published in the GitHub repository vindm/dotclaude (1 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,495 tokens. 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 2026-08-31.
Other agents, from other repositories
css-expert
Expert in modern CSS including CSS Grid, Flexbox, custom properties, container queries, CSS layers, native nesting, modern color spaces, and advanced styling techniques. Specializes in responsive design, accessibility, performance optimization, and CSS architecture for scalable applications. Examples: Context: User…
config-safety-reviewer
Configuration safety specialist focusing on production reliability, magic numbers, pool sizes, timeouts, and connection limits. Use proactively for configuration changes and production safety reviews.
design-md-consumer
Load when the user's project ships a design contract — a .design file (schema design.v1) or a DESIGN.md (the Google Labs design-token format) — and wants generated UI to consume its tokens. Discovers the contract, resolves precedence, normalizes the tokens, and produces TypeScript / CSS variables / a Tailwind config /…
agentation-fix-loop
Load when the user is running an Agentation two-session workflow and you are the Session 2 (fix) side. Watches MCP annotations from the critique session, reads each annotation's selector + file path + critique, applies the fix in code, commits, and loops to the next. Use only when the agentation MCP server is…
prompt-ux-designer
Designs the prompt input experience for AI features — input pattern, suggestion system, feedback loop, history, and constraints. Use before implementing any AI prompt input or chat interface.
design-auditor
Master audit orchestrator. Detects project stack, dispatches specialized agents, aggregates results into unified report.