reimagine-it-audit — Claude Code skill: what it does, how to install
Runs 19 deterministic checks on generated HTML to catch layout, typography, and accessibility regressions before shipping. Needs no LLM or API keys.
npx skills add Kayforkind/reimagine-it --skill audit -a claude-codeWhat reimagine-it-audit does
The skill evaluates HTML files against 19 rules divided into six groups: typography, color palette, motion, content sanity, structure, and performance. It identifies banned default fonts, text lines wider than 65ch, transitions on unoptimized properties, unstyled selection pseudo-elements, missing image dimensions, and placeholder copy.
Execution happens locally via npx reimagine-it audit or the CI mirror scripts/audit.py. The process never queries an LLM, keeping runs fast and free of external API tokens. For build pipelines, it supports a --json flag and standard exit codes: 0 for clean output, 1 for warnings, and 2 for blocking failures.
Rules MOT-02 and MOT-03 serve as hard blockers. Stripping focus indicators or omitting prefers-reduced-motion queries will fail the audit immediately.
What it helps with
- catch missing focus-visible rings and stripped outline styles
- detect placeholder copy, emoji clutter, and stacked br tags
- flag images missing explicit dimensions and heavy pages without content-visibility
- enforce typographic line length and prevent default font fallbacks
Example prompts for the agent
- /reimagine-it audit path/to/page.html
- Audit gold/webpage/after.html with the --verbose flag
- Run /reimagine-it audit --json to check design quality in CI
How to install reimagine-it-audit
1Skills CLI (Vercel)
npx skills add Kayforkind/reimagine-it --skill audit -a claude-code2Manually: put the files in place
~/.claude/skills/audit/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/audit/
1Skills CLI (Vercel)
npx skills add Kayforkind/reimagine-it --skill audit -a codex2Manually: put the files in place
.agents/skills/audit/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add Kayforkind/reimagine-it --skill audit -a cursor2Manually: put the files in place
.cursor/skills/audit/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install Kayforkind/reimagine-it audit2Manually: put the files in place
.github/skills/audit/SKILL.md1Skills CLI (Vercel)
npx skills add Kayforkind/reimagine-it --skill audit -a gemini-cli2Platform CLI
gemini skills install https://github.com/Kayforkind/reimagine-it --path skills/reimagine-it/auditManually: put the files in place
.devin/skills/audit/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add Kayforkind/reimagine-it --skill audit -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about reimagine-it-audit
How to install reimagine-it-audit in Claude Code?
Run npx skills add Kayforkind/reimagine-it --skill audit -a claude-code. The skill writes to ~/.claude/skills/audit/SKILL.md.
What does reimagine-it-audit do?
It runs 19 deterministic quality checks across typography, palette, motion, content, DOM structure, and rendering performance.
Does reimagine-it-audit require an API key or LLM calls?
No. The audit relies purely on local Node.js or Python scripts without invoking any language model.
Source text
SKILL.md · MIT
---
name: reimagine-it-audit
description: >-
Design Health — runs 19 deterministic quality checks on HTML output.
Use when the user says /reimagine-it audit, "audit this page", "check design quality",
or wants to verify craft-floor compliance before shipping. Catches blank plates,
missing focus rings, non-compositor motion, off-palette accents, dead motion,
fabricated content. No LLM, no API key. CI-ready with --json and exit codes.
Part of the reimagine-it Content-Derived Design suite.
license: MIT
metadata:
author: Kayforkind
version: "2.15.0"
parent: reimagine-it
hosts:
- claude-code
- cursor
- codex
- copilot
- gemini-cli
keywords:
- audit
- quality-check
- craft-floor
- design-health
- html-lint
- accessibility-check
- content-derived-design
category: Testing
capabilities:
- html-audit
- quality-verification
- craft-floor-enforcement
trigger_phrases:
- /reimagine-it audit
- audit this page
- check design quality
- verify craft floor
- design health check
- audit my HTML
---
# /reimagine-it audit
**Parent:** [../SKILL.md](../SKILL.md) — the full Content-Derived Design engine. This sub-skill handles only the deterministic quality check.
Runs `npx reimagine-it audit` — 19 checks across 6 categories (`src/audit.js`; Python `scripts/audit.py` is the CI mirror):
| Category | Checks | What it catches |
|----------|--------|-----------------|
| Typography | 3 | Banned default fonts, type-scale range, measure > 65ch |
| Palette | 3 | Palette drift, `transition: all`, un-styled `::selection` |
| Motion | 4 | Outline removal, missing `prefers-reduced-motion`, missing `:focus-visible`, non-compositor animation |
| Content | 4 | Placeholder copy, vibe adjectives, emoji farm, `<br><br>` spacing |
| Structure | 3 | CDN, external font fetch, figure system doing no work |
| Performance | 2 | Images without dimensions, long pages without `content-visibility` |
## Usage
```bash
# Agent
/reimagine-it audit path/to/page.html
# CLI (Node — no Python required)
npx reimagine-it audit gold/webpage/after.html
npx reimagine-it audit gold/webpage/after.html --verbose
npx reimagine-it audit gold/webpage/after.html --json
# CI mirror
python scripts/audit_all.py
```
## Exit codes
| Code | Meaning |
|------|---------|
| 0 | CLEAN — no issues |
| 1 | WARNINGS — advisory only, no blockers |
| 2 | FAIL — must fix before shipping |
## Procedure
1. If the user passed a file path, run `python scripts/audit.py <path>`.
2. If no path, run `python scripts/audit_all.py` to audit all gold files.
3. Report the verdict with a summary table.
4. For each failure, name the specific rule and the fix.
5. If `--verbose`, show per-category breakdown.
6. If `--json`, output machine-readable JSON for CI.
## Must not
- Use an LLM for the check — this is deterministic Python only
- Skip the craft-floor rules (MOT-02, MOT-03 are hard blockers)
- Report "clean" without actually running the script
- Invent checks not in `scripts/audit.py`
© Kayforkind · MIT · Kayforkind/reimagine-it