Front End Interview Handbook: questions/ plus website/, one root script, translations that start mid-corpus

The root package.json has one script, and it is not the one you need

The root manifest is unusually thin. It is marked private, it declares packageManager [email protected], it has a single script, prepare running vp config, and its only devDependencies are typescript at ^5.8.2 and vite-plus referenced as catalog:. Consequence: cloning the repository and running pnpm install at the root executes one configuration step and nothing else. There is no build script, no dev script, and no test script at the top level, so the commands a new contributor needs live in per-package manifests inside the pnpm workspace, and the README documents none of them. The catalog: reference is the other thing to understand, because it means the vite-plus version is resolved through the workspace catalog rather than pinned in this file, which is a sensible monorepo layout and a poor first-run experience for anyone expecting a documented command in the README.

questions/ and website/ split the corpus from the rendering, and Travis still sits at the top level

The tree has two content roots, questions/ and website/, plus assets/, a vite.config.ts, a .vite-hooks/ directory, and an .editorconfig. Consequence: this is a Vite site with content in one place and rendering in another, and the hooks directory suggests build-time steps that validate or transform content rather than hand-authored pages. For someone who spots a wrong answer, that means editing content under questions/ and waiting for a deploy, rather than changing a document and seeing a diff. The repository also carries a .travis.yml at the top level next to the .github directory, so two CI configurations are present, one of them from a service that predates GitHub Actions. Alongside them sit pnpm-workspace.yaml and pnpm-lock.yaml, which is where the actual dependency graph lives given how little the root manifest declares.

All eight translations link to the same JavaScript questions page, not to a translated index

The translations list has eight entries. English points at the site root. The other seven, Simplified Chinese, Japanese, Korean, Polish, Portuguese, Russian and Tagalog, all point at a locale path that ends in javascript-questions/. Consequence: a reader who follows the Tagalog link lands inside one section of the corpus rather than on a translated table of contents, and the list contains no link at all for the system design page or the company questions page that the English index does cover. So the translation list looks complete at eight languages and is partial in coverage, and the way it is written does not show that. The mitigating detail is in the contributing guide, which explicitly covers translating the content, so this is an invited gap rather than an oversight, and it is also the clearest place a new contributor can add value.

The whole premise is that front end interviews are not algorithm interviews

The framing sentence is the reason this project exists separately from a general preparation guide: unlike typical software engineer job interviews, front end job interviews have less emphasis on algorithms and have more questions on detailed knowledge and expertise about the domain, HTML, CSS and JavaScript to name a few areas. Consequence: the split is by discipline rather than by seniority, and the project's own answer to a process that screens on algorithms is a different handbook. The README points to the Tech Interview Handbook for general coding interviews, algorithms, behavioral questions, and an interview cheatsheet, and describes it as what to read if you are looking for generic interview preparation. The two are complementary rather than substitutes, and the illustration credit for that sister site goes to unDraw.

The trivia section is anchored to someone else's question list, so its scope is that list

One of the five content areas is described as answers to the h5bp Front-end Developer Interview Questions. That single clause tells you how the section is built. Consequence: the trivia material is reactive rather than curated from a syllabus, so its coverage is exactly the upstream list's coverage and it ages at whatever rate that list ages. It is a good on-ramp if you have never seen those questions before, and a poor instrument for finding gaps in your own knowledge, because the index cannot tell you what is not on the list. The other four areas are organised the opposite way, by interview format, by system design topic, by language area covering JavaScript, DOM and algorithms, and by company, which means the index tells you what exists without telling you how deep any of it goes.

Practice is a separate paid product, and the README doubles as its storefront

After the free content, the README points at GreatFrontEnd, described as a platform built by the author, offering more than 200 practice questions each with multiple solutions from senior front end engineers, plus automated test case suites to help identify what is wrong with your code. The named resources are study plans, focus areas, preparation by stage, and individual framework questions, and a separate block advertises a limited-time 20% discount on a lifetime plan whose reference solutions are written by ex-FAANG interviewers. The author's social accounts, a buymeacoffee link, and a pointer to Lago, the author's data structures library, sit in the same document. Consequence: the repository is MIT licensed and free, and the README is at once the best free resource and the funnel to a paid product, so the free handbook's scope stops where the paid practice begins. Lago is explicitly labelled as meant for reference and studying, not for production use.

Distribution is a website only, with no release and nothing to pin

The project publishes eight language versions as web pages and holds no downloadable artefact. The repository has no GitHub releases, and the root package is marked private so nothing is published to a registry either. Consequence: a reader who wants the handbook offline has to collect the pages themselves, and a reader who wants to cite a fixed version has nothing to pin, because the site's content changes whenever the branch changes and the last push on the default branch was 2026-08-13. That single date is the only freshness signal available, since there is no version history to read. It also explains why this handbook is regularly searched for in a downloadable form while the project itself only offers a site, which is a distribution mismatch rather than a content one. The licence line is clear in comparison, stating that all projects and packages in the repository are MIT licensed.

Editorial conclusion

The Front End Interview Handbook fits a reader who is preparing for a process that emphasises front end domain knowledge over algorithms, who reads English or one of the seven translations, and who is willing to read on a website rather than download anything. It does not fit a reader who needs a fixed version to cite, a downloadable copy, or a translated table of contents, and it does not cover the algorithmic half of an interview loop. Before you plan around it, check five things: whether the content pages for your language exist beyond the JavaScript questions page the translation links point at; whether the system design and company question sections have depth for the process you are facing, since the index shows what exists but not how far it goes; whether you also need the separate Tech Interview Handbook for algorithms and behavioral material, which the README recommends for general preparation; how you will run the site, since the root package.json has only a prepare script and the rest lives in workspace packages the README never mentions; and what changed since 2026-08-13, which is the last push on the default branch and the only timestamp you have for judging freshness.

Frequently asked questions

Official sources

1. Official documentation
2. Official README
3. Project repository