nextarter-chakra: a Next.js starter with the tooling decisions already made

A template that ships with the stack chosen

nextarter-chakra is described as a battery packed template, or boilerplate, to initialise a Next.js app with a Chakra UI and TypeScript setup, and the README frames it as a Next.js project bootstrapped with create-next-app, added with Chakra UI and TypeScript setup, with the instruction to start developing right away. That is the whole proposition. The alternative to this template is not writing code, it is making the boring decisions first, and the list of those decisions in the feature section is the actual product. The repository also ships deploy buttons for Vercel, Netlify and Railway, plus a StackBlitz link, so you can see the thing running before you have cloned anything, and there is a running instance at nextarter-chakra.sznm.dev. It is reasonable to say plainly what this is not, though. It is not a component library and it is not a framework, it is a starting point that a team will either keep or throw away, and the second outcome is common.

The stack it commits you to

The feature list names the framework as Next.js 16 with the App Router, the UI library as Chakra UI v3, the styling engine as Emotion through Chakra UI together with next-themes, the language as TypeScript, the linting and formatting as Biome, end to end testing as Playwright, the build system as Turborepo, and the icons as React Icons. Each of those is a real decision with a cost. Turborepo in particular is a monorepo tool in a project that starts as one application, which is sensible if the template is expected to grow into a workspace, and unnecessary if it will not. Biome replaces the ESLint and Prettier pairing, which is fewer moving parts but a younger toolchain. The manifest shows how specific the choices are, with next pinned at 16.3.1, react and react-dom at 19.2.8, @chakra-ui/react at ^3.37.0, @biomejs/biome at 2.5.9, @playwright/test at ^1.63.0 and ultracite at 7.10.6, and the package manager declared as [email protected].

The split layer structure and why it is drawn out

The system architecture section is the part to understand before you copy the repository, because it is a position rather than a convention. The project uses a split layer architecture to separate Next.js routing concerns from React UI logic, and the diagram makes the flow explicit. The routing layer lives in src/app, where layout.tsx feeds the providers and page.ts holds the logic. The logic layer lives in src/lib, where the providers lead to a layout component with a header and a footer, and the logic leads to a home page component. The UI layer in src/components holds the shared components that all of it renders. The tree is the same thing in a form you can read:

src/
├── app/            # Next.js App Router (Entry points)
│   ├── layout.tsx  # Global RootLayout & Providers
│   └── page.ts     # Re-exports Home from lib/pages
├── components/     # Shared UI components
│   └── ui/         # Primitives (Buttons, Inputs, etc.)
└── lib/            # Application Business Logic
    ├── layout/     # Shell components (Header, Footer)
    ├── pages/      # Page implementations (Home, etc.)
    └── styles/     # Theme configuration

The practical effect is that a page route re-exports an implementation rather than containing it, which keeps the App Router thin and leaves the React code testable and portable.

Getting started in four commands

The project uses pnpm, and the getting started section is short enough to quote. First the dependencies:

pnpm install

Then the development server, which is next dev:

pnpm dev

Open http://localhost:3000 to view it in the browser. For a production build:

pnpm build

And the end to end tests, which run Playwright:

pnpm test:e2e

The build step has a postbuild hook wired into it, and the package manifest shows that the postbuild script runs next-sitemap with its own configuration file, so a static site map is generated as part of building the site rather than as a separate step. There is also a Turbo variant, build:turbo, which runs the build through the task runner across the workspace, and a type check script that runs tsc with no emit.

Quality gates, hooks and the release path

The code quality section names Biome as the formatter and linter, with two commands:

# Check code issues
pnpm biome:check

# Fix issues automatically
pnpm biome:fix

The manifest goes further than the README does. There is a biome:ci script that runs Biome in CI mode, an ultracite:check and ultracite:fix pair, a type:check script that runs tsc with no emit, and a check:turbo script that runs the ultracite check and the type check together through Turborepo. The Turbo binary is pinned in the scripts with pnpm dlx [email protected], which keeps the task runner version stable for everyone. Around the checks sit the commit machinery, with husky installed through the prepare script, a commitlint configuration using the conventional config, a commitizen cz configuration in .czrc, lint-staged configured in .lintstagedrc.json, and knip for finding unused code. Releases are cut with commit-and-tag-version, run through cross-env, and the push-release script pushes the tags to origin main.

A template repository with more going on than usual

The top level of the repository is busier than a typical starter. Alongside the expected src, public, e2e, next.config.ts, turbo.json and tsconfig.json there are several deployment configurations, with vercel.json, netlify.toml and wrangler.toml, and a next-sitemap.config.js for the search engine output. Then there is the tooling and documentation layer. The tree carries an AGENTS.md, a CONTRIBUTING.md, a SPEC.md, a WARP.md and a references.md, plus an opencode.jsonc configuration and a skills-lock.json, and there are editor and agent specific directories, .agents, .claude, .codegraph, .cursor and .zed, together with .husky for the git hooks and .versionrc for the versioning tool. A .czrc, .npmrc and .nvmrc round out the configuration, and renovate.json is present for dependency updates. Read that as a signal about how the project expects to be worked on, because a template that ships agent instructions and a spec document is being maintained as a product rather than shipped once and forgotten.

Versions, licence and the mismatch in the docs

The package metadata is where the practical details live. The package is named nextarter-chakra, is private, and sits at version 0.1.0. The runtime dependencies are @chakra-ui/react, @emotion/react, commit-and-tag-version, knip, next at 16.3.1, next-themes, react and react-dom at 19.2.8 and react-icons, and the development set brings in the Biome and Playwright packages, the Chakra UI CLI, the commitlint family, commitizen, cross-env, husky, lint-staged, next-sitemap and typescript. The engines field asks for node ^24.11.x, and the devEngines field asks for node ^24.19.0 with an onFail of download, so a machine with the wrong Node will be asked to fetch the right one. The licence is MIT, there are no GitHub releases, the last push was on 2026-10-22 and the repository is not archived. One discrepancy deserves a flag before you copy it, since the README lists the language as TypeScript 5 while the manifest pins typescript to about 7.0.2, which tells you the documentation lags the dependency set.

Editorial conclusion

nextarter-chakra fits a developer who wants to start a React application without a week of setup, and a team that has agreed on Biome, Playwright and Turborepo and wants a project that already reflects that. It is the wrong choice if you do not want Chakra UI, since the component library is baked into the template rather than layered on top of it, and it is a poor fit if your target runtime is not a recent Node, because the package metadata pins node ^24.11.x with a development engine of ^24.19.0. Before you start, read the split layer structure and the repository scripts so you know what the template will hand you, and check the versions in package.json against the ones the README lists, because the two already disagree about TypeScript.

Frequently asked questions

Official sources

1. agustinusnathaniel/nextarter-chakra on GitHub
2. Issues
3. License: MIT
4. Project website
5. README