--- title: Next js Is AI Native Now My 10 Part Setup For AI Native... description: Next.js 16.2 and 16.3 introduce a series of AI-native features designed to improve coding agent workflows. Key additions include an agents.md file bundled with... canonical: https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Next js Is AI Native Now My 10 Part Setup For AI Native Workflow | daily.dev og:description: Next.js 16.2 and 16.3 introduce a series of AI-native features designed to improve coding agent workflows. Key additions include an agents.md file bundled with... og:url: https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu og:image: https://api.daily.dev/og/posts/9cs4TxBZu.png og:image:alt: Next js Is AI Native Now My 10 Part Setup For AI Native Workflow og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Next js Is AI Native Now My 10 Part Setup For AI Native Workflow **[ByteGrad](https://daily.dev/sources/bytegrad)** · 25 min read · 1 upvotes · 0 comments ## Summary Next.js 16.2 and 16.3 introduce a series of AI-native features designed to improve coding agent workflows. Key additions include an agents.md file bundled with new projects that points agents to version-specific Next.js docs, significantly boosting success rates across models. Browser log forwarding lets agents see client-side errors without manual copy-pasting. A dev server lock file prevents agents from spinning up duplicate servers. New first-party skills — especially the 'next dev loop' skill — give agents a full feedback loop: they can drive a real Chrome browser, inspect the React component tree, read server logs, and verify runtime behavior after edits. The Gemini K3 model currently tops the Next.js evals leaderboard. Supabase is highlighted as a complementary tool with its own MCP server and agent skill that inlines critical security requirements (like RLS policies) directly into the skill file so agents can't skip them. The combination of Next.js skills, Supabase MCP, and agent browser enables building and end-to-end testing full-stack apps autonomously. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the agents.md file in Next.js 16.2 and what does it do? Agents.md is a file automatically created when you run create-next-app that bundles the Next.js documentation directly into the project so coding agents like Codex or Claude Code can reference it without relying on outdated training data. It significantly improves benchmark success rates on the Next.js Evals, pushing many models from around 75% to about 96% accuracy, since agents no longer skip fetching current docs. _Developers tuning their AI coding setup for Next.js can track workflow changes like this through daily.dev._ ### What is the agent browser package in Next.js and how is it different from next-browser? Agent browser is the successor to the next-browser package introduced in Next.js 16.2, now expanded to work outside Next.js projects too. It lets a coding agent drive a real Chrome browser, inspecting the DOM, console, network requests, web vitals, and the React component tree with React DevTools-level introspection, giving agents runtime visibility beyond just reading source files. _Anyone wiring agent tooling into a browser-based workflow can follow developments like agent browser via daily.dev._ ### How does Supabase's new agent skill improve RLS security when using AI coding agents? Supabase's skill.md inlines critical row-level security requirements directly into the skill file, roughly 100 lines, so the agent reads them the moment the skill loads rather than needing to fetch reference docs on demand, which agents sometimes skip. This addresses agents previously skipping RLS policies, hallucinating CLI commands, or creating views that silently bypass RLS due to outdated training data. _Teams worried about agents mishandling Supabase RLS can keep up with security-focused tooling changes on daily.dev._ ## Similar posts on daily.dev - [Next.js 16.3: AI Improvements](https://daily.dev/posts/next-js-16-3-ai-improvements-cq3eiij4w) · Next.js · 34 upvotes · 1 comments - [Next.js 16.2: 400% Faster Dev Startup, Faster Rendering, and Deeper Tooling for AI Agents](https://daily.dev/posts/next-js-16-2-400-faster-dev-startup-faster-rendering-and-deeper-tooling-for-ai-agents-vt3mgjvy9) · InfoQ · 74 upvotes · 12 comments - [Building Next.js for an agentic future](https://daily.dev/posts/building-next-js-for-an-agentic-future-ucotwnjl7) · Next.js · 31 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#nextjs](https://daily.dev/tags/nextjs), [#mcp](https://daily.dev/tags/mcp), [#claude-code](https://daily.dev/tags/claude-code), [#supabase](https://daily.dev/tags/supabase) [View this post on daily.dev](https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Next js Is AI Native Now My 10 Part Setup For AI Native Workflow","url":"https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu"},"datePublished":"2026-07-25T07:07:23.290Z","dateModified":"2026-09-14T07:51:54.573Z","description":"Next.js 16.2 and 16.3 introduce a series of AI-native features designed to improve coding agent workflows. Key additions include an agents.md file bundled with...","image":"https://i.ytimg.com/vi/6jFHRIUTRHo/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/6jFHRIUTRHo/sddefault.jpg","isAccessibleForFree":true,"articleSection":"ByteGrad","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"ByteGrad","logo":"https://media.daily.dev/image/upload/s--UnJ6vKLR--/f_auto/v1732552127/logos/bytegrad","url":"https://daily.dev/sources/bytegrad"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,nextjs,mcp,claude-code,supabase","timeRequired":"PT25M","video":{"@type":"VideoObject","name":"Next js Is AI Native Now My 10 Part Setup For AI Native Workflow","description":"Next.js 16.2 and 16.3 introduce a series of AI-native features designed to improve coding agent workflows. Key additions include an agents.md file bundled with...","thumbnailUrl":"https://i.ytimg.com/vi/6jFHRIUTRHo/sddefault.jpg","uploadDate":"2026-07-25T07:07:23.290Z","duration":"PT25M","url":"https://api.daily.dev/r/9cs4TxBZu","embedUrl":"https://www.youtube.com/embed/6jFHRIUTRHo"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"ByteGrad","item":"https://daily.dev/sources/bytegrad"},{"@type":"ListItem","position":3,"name":"Next js Is AI Native Now My 10 Part Setup For AI Native Workflow"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/next-js-is-ai-native-now-my-10-part-setup-for-ai-native-workflow-9cs4txbzu#faq","mainEntity":[{"@type":"Question","name":"What is the agents.md file in Next.js 16.2 and what does it do?","acceptedAnswer":{"@type":"Answer","text":"Agents.md is a file automatically created when you run create-next-app that bundles the Next.js documentation directly into the project so coding agents like Codex or Claude Code can reference it without relying on outdated training data. It significantly improves benchmark success rates on the Next.js Evals, pushing many models from around 75% to about 96% accuracy, since agents no longer skip fetching current docs. Developers tuning their AI coding setup for Next.js can track workflow changes like this through daily.dev."}},{"@type":"Question","name":"What is the agent browser package in Next.js and how is it different from next-browser?","acceptedAnswer":{"@type":"Answer","text":"Agent browser is the successor to the next-browser package introduced in Next.js 16.2, now expanded to work outside Next.js projects too. It lets a coding agent drive a real Chrome browser, inspecting the DOM, console, network requests, web vitals, and the React component tree with React DevTools-level introspection, giving agents runtime visibility beyond just reading source files. Anyone wiring agent tooling into a browser-based workflow can follow developments like agent browser via daily.dev."}},{"@type":"Question","name":"How does Supabase's new agent skill improve RLS security when using AI coding agents?","acceptedAnswer":{"@type":"Answer","text":"Supabase's skill.md inlines critical row-level security requirements directly into the skill file, roughly 100 lines, so the agent reads them the moment the skill loads rather than needing to fetch reference docs on demand, which agents sometimes skip. This addresses agents previously skipping RLS policies, hallucinating CLI commands, or creating views that silently bypass RLS due to outdated training data. Teams worried about agents mishandling Supabase RLS can keep up with security-focused tooling changes on daily.dev."}}]} ```