Skip to content
← Tools / Skills
Skill · .md packageVercel/10 trustVerified

Vercel Agent Skills

Vercel's official engineering rule packs — React/Next.js performance, web-interface design guidelines, and cost/perf audits.

Visit official site →

About

Vercel's official collection of Agent Skills for AI coding agents, following the agentskills.io format. Ships 8 skills: react-best-practices (40+ Vercel Engineering rules), web-design-guidelines (100+ a11y/perf/UX audit rules), vercel-optimize (cost/perf audit of deployed Vercel projects), plus writing guidelines, React Native guidelines, view transitions, composition patterns, and claimable Vercel deploys.

What it can do

  • React/Next.js performance rules (waterfalls, bundle size, SSR, re-render optimization)
  • 100+ web-interface audit rules (a11y, focus states, forms, animation, typography, images, dark mode)
  • Vercel project audits (cost, caching, function usage, billing)
  • Claimable instant deployments via vercel-deploy-claimable
  • Writing guidelines, React Native guidelines, view transitions, composition patterns

Setup

  1. npx skills add vercel-labs/agent-skills (all skills) or npx skills add vercel-labs/agent-skills --skill <name> (e.g. --skill react-best-practices, --skill web-design-guidelines)
  2. Skills auto-activate when relevant tasks are detected.

Prerequisites

Any agent that loads SKILL.md files; for vercel-optimize, a deployed Vercel project.

Use cases

codedesigndevops

Works with

Any agent (.md)Claude CodeCursorCodex

Trust score — 9/10

/10 trustVerified

Official Vercel Labs repo, MIT, 100K+ installs per top skill on skills.sh, actively maintained (275 commits), docs live in the repo README. Markdown-only guidance (no executable code), so minimal risk surface. Slightly narrower adoption than the giant community repos, but official and current.

★ 32,292 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
8.4/10
Maintenance · recency, releases, issue load
7.6/10
Openness · license + docs
10.0/10
Manual review · sanity read + red-flag check
9.0/10
  • · 32,292 GitHub stars (log-scaled vs catalog peers)
  • · last push 45d ago; 3 releases/12mo; 180 open issues
  • · manual review: narrower adoption; skills are thin wrappers

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.