[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f37sfqvouh0tc0":3},{"code":4,"message":5,"data":6,"timestamp":32},10000,"success",{"full_name":7,"position":8,"description":9,"language":10,"stars":11,"forks":12,"stars_gained":13,"age_days":14,"hn_threads":15,"one_liner":22,"summary_md":23,"topic":24,"github_topics":27,"license":28,"homepage":29,"updated_at":30,"history":31},"vercel-labs\u002Fagent-skills",15,"Vercel's official collection of agent skills","JavaScript",32947,2880,1494,320,[16],{"id":17,"title":18,"points":19,"comments":20,"posted_at":21},46623875,"Agent Skills: AI Agents for React and Next.js Workflows",2,0,"2026-01-14","Packaged skill instructions that teach AI coding agents Vercel's React, Next.js and deployment conventions.","## The problem\n\nAI coding agents know React roughly, but they do not know your team's performance rules, accessibility checklist, writing handbook or deploy flow. Without that context they produce code that passes a glance and fails review: waterfalls, oversized bundles, missing aria labels, docs in the wrong voice. This repo packages those conventions as installable skills so the agent loads the right guidance when a relevant task shows up.\n\n## What's different\n\nMost teams solve this with a rules file in the repo or a long system prompt, which either leaks into every conversation or gets ignored once it grows. Here the guidance is split into named skills with trigger descriptions, so context is loaded on demand. The content is also opinionated rather than generic: the React rules are ranked by impact (waterfalls and bundle size first, micro-optimizations last), and the writing skill encodes Vercel's own handbook, including its banned-word list. The deploy skill is the odd one out, reaching outside the editor to package and upload a project. That is useful for chat-based agents, and it is also where the collection becomes a funnel toward Vercel's platform.\n\n## Who it's for\n\nTeams on Next.js, React or React Native who use Claude Code, Cursor or similar agents and want reviews and generated code to match a house style. It is strongest for frontend platform and DX people who would otherwise maintain this checklist themselves. It is a poor fit if you are on another framework, if you already have a mature internal rules set, or if you are not on Vercel and need deploy tooling: the optimization skill reads Vercel metrics and has little to offer elsewhere.\n\n## Cost of adoption\n\nInstallation is one command, `npx skills add vercel-labs\u002Fagent-skills`, and the artifacts are plain Markdown plus optional scripts, so you can read and fork rules. The README declares MIT while the repo metadata lists no license, so confirm before vendoring. Skills change behaviour silently once installed; there is no evaluation harness here to tell you whether a rule helped or hurt. Expect periodic sync work as the upstream rules evolve, and expect the deploy skill to require Vercel access.\n\n## Verdict\n\nThis is a curated opinionated checklist from a vendor that ships Next.js, not a framework, and its value tracks how close your stack is to Vercel's defaults. The repository is young, created in December 2025, yet already at 32.9k stars with +1,494 in the tracked month, which looks like fast adoption rather than a long-established staple; there is no daily trending history and Hacker News interest is thin. Treat it as a useful starting point to copy into your own conventions, and read each rule before trusting it in CI.",{"slug":25,"name":26},"developer-tools","Developer Tools",[],null,"https:\u002F\u002Fskills.sh\u002Fvercel-labs\u002Fagent-skills","2026-10-24T18:05:57Z",[],1792879654]