julyssaalexandrauriarteanccota/inventori/web-design-guidelines
- community
Checking that the SKILL.md file is still in the repository.
Audit UI code against Vercel Web Interface Guidelines for design, accessibility, and UX compliance. Use when: reviewing frontend PRs, auditing pages/components, and generating terse file:line findings.
Install
npx vskill@latest install julyssaalexandrauriarteanccota/inventori/web-design-guidelinesSKILL.md
Web Design Guidelines Audit
Objective
Audit UI code against Vercel Web Interface Guidelines with concise, actionable findings that are easy to remediate.
When To Use
- Reviewing frontend pull requests before merge.
- Auditing pages, layouts, and components for visual and interaction quality.
- Catching accessibility, UX, and consistency regressions in production-facing UI.
Inputs
- One or more file paths, folders, or glob patterns.
- Optional context: product area, target device, design system constraints.
If no files are provided, ask the user which files to review before continuing.
Procedure
- Refresh the source of truth before each audit.
- Fetch the latest Vercel guidelines from the remote repository before reviewing code.
- If the primary remote document is unavailable, use the best available official fallback from the same repository and state that fallback in output.
- Resolve scope and files.
- If arguments include explicit files, audit only those files.
- If arguments include globs, expand and audit matching UI files.
- If arguments are empty, prompt for files or patterns.
- Run a category-based review.
- Design: typography scale, spacing rhythm, color contrast, hierarchy, visual clarity.
- Accessibility: semantics, labels, keyboard flows, focus visibility, ARIA misuse, readable states.
- UX: affordance clarity, feedback states, loading/empty/error states, responsiveness.
- Produce terse findings for fast remediation.
- Output each issue on one line in this format:
path/to/file.ext:line | severity | rule | short fix
- Keep severity to: high, medium, low.
- Do not append narrative summaries or extra prose.
- If no issue is found, return: No findings.
- Add completion checks.
- Confirm every finding includes an exact file and line.
- Remove duplicate findings that describe the same root issue.
- Prioritize high-impact issues first, then medium, then low.
Decision Points
- If guidelines cannot be fetched: continue with fallback and explicitly flag reduced confidence.
- If requested files are missing: report unresolved paths and request corrected targets.
- If generated findings are verbose: compress each finding to one line while preserving actionability.
Quality Criteria
- Uses latest available remote guidelines at review time.
- Findings are terse, scannable, and directly mappable to edits.
- Covers design, accessibility, and UX in every audit.
- Avoids generic advice; each finding points to concrete code.
Output Example
apps/web/src/app/(public)/page.tsx:42 | high | Missing visible focus state on primary CTA | Add :focus-visible style with >= 3:1 contrast ring. apps/web/src/components/forms/contact-form.tsx:88 | medium | Input lacks associated label | Add <label htmlFor> and matching id.
Example Prompts
- /web-design-guidelines apps/web/src/app/**
- /web-design-guidelines apps/web/src/components/**/*.tsx
- /web-design-guidelines Audit checkout and support pages for accessibility and UX regressions.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryChecking.github/skills/web-design-guidelines/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/julyssaalexandrauriarteanccota/inventori/web-design-guidelines)