Ebazhanov/linkedin-skill-assessments-quizzes: What This Repository Actually Contains Now

What the repository is for, and who it is actually aimed at

LinkedIn skill assessments were short multiple-choice tests attached to a profile. LinkedIn stopped offering them, and the README opens with that fact: "As of December 2023 skills assessments are no longer available on LinkedIn, but you can still master your knowledge here." So the repository is no longer a way to pass a test you can still take. It is a study corpus.

The audience follows from that. Someone preparing for a technical interview in JavaScript, React, Git, HTML, jQuery, MongoDB, Java, Go or Python can work through the matching directory as a self-check. Someone who wants to see how a topic is commonly framed in multiple-choice form, including the wrong answers, gets that too, because each quiz file lists options rather than just answers. Contributors who want to add explanations or translations are the third group, and the README points them at a separate repository for the source code behind the content.

What it is not is a certification path. Nothing here is issued by LinkedIn, and the disclaimer states the owners accept no liability for how the content is used.

Directory-per-topic layout and the markdown question format

The repository is a flat set of topic directories at the top level: accounting/, adobe-photoshop/, aws-lambda/, aws/, bash/, css/, cybersecurity/, django/, git/, go/, html/, java/, javascript/, jquery/, json/, kotlin/, and many more in the same pattern. Each directory holds one or more markdown files, typically named after the topic, for example adobe-photoshop/adobe-photoshop-quiz.md. Translations sit beside the English file with a language suffix, so accounting/ contains accounting-quiz-ch.md, accounting-quiz-es.md, accounting-quiz-fr.md and accounting-quiz-it.md alongside accounting-quiz.md.

The README's table of contents is the index. Each row carries a topic, a Passed/Failed column, a Translation column with flag images linking to the localised files, a Questions count, an Answers count, and a list of GitHub handles for people who review changes to that topic. Accounting, for instance, is listed with 75 questions and 75 answers, and its Passed/Failed cell reads "needs updating". That column is the closest thing to a quality signal in the whole project, and it is worth reading before trusting any single file.

The architecture is deliberately boring: markdown in, markdown out. There is no database, no API and no build step for the content itself. That is why the whole thing can be consumed by a static site generator (the repository carries a _config.yml and a GitHub Pages homepage) without any transformation layer.

Reading a quiz file locally and running the formatter

There is no package to install for the content. The README gives no install steps for the question bank itself; you clone the repository and open the files. The only tooling declared in the repository is a formatter. The package.json defines a single script, format, which runs Prettier over markdown, and Prettier is listed as a devDependency at ^3.6.2. The script itself is declared in package.json as prettier --write "**/*.md", so it rewrites every markdown file in the tree. Expect a large diff on a fresh clone if the files were last formatted with a different Prettier version. That is a formatting pass, not a content check; it will not tell you whether an answer is right.

To read a single topic without a browser, open the file directly. The README links individual quiz files through github1s.com and github.dev, for example adobe-photoshop/adobe-photoshop-quiz.md, which is the fastest route if you only want to skim one topic. The README additionally lists three practice front ends, including MD2Practice and a terminal tool called Kodyfire, for people who would rather not read raw markdown.

The repository layout is the whole delivery mechanism here. There is no index page generated from the content, no search, and no per-topic metadata beyond what the README table records by hand. If you want to find every question mentioning a particular keyword, you are grepping the markdown tree yourself.

The Passed/Failed column is the honest part of this project

Most answer banks present themselves as complete. This one does not. The table of contents has a Passed/Failed column, and at least one entry, Accounting, is explicitly marked "needs updating" while still listing 75 questions and 75 answers. That combination is the useful signal: a topic can look fully populated and still be flagged as stale.

The practical consequence is that a topic directory existing tells you almost nothing. The Questions and Answers counts in the table are the only per-topic completeness figures the README publishes, and they are maintained by hand. If a row shows a mismatch, or if the Passed/Failed cell carries a warning, treat the file as a draft.

There is also a structural risk that comes with the format itself. Multiple-choice questions about tools change when the tools change. A Git quiz answer written against one default branch name or one command flag can quietly become wrong, and Prettier will happily reformat it without noticing. The repository's review model is the mitigation: each row names the people who review pull requests for that topic, and the README asks contributors to add an explanation or reference link to every answer. That request is a quality mechanism, not a guarantee, and it depends on reviewers being active for that directory.

Licence: AGPL-3.0 in the repository, MIT in package.json

The repository's LICENSE file is AGPL-3.0. The package.json, however, declares "license": "MIT". Those two statements do not agree, and anyone reusing the content in a product should resolve that discrepancy rather than pick whichever is convenient.

AGPL-3.0 is a copyleft licence with a network clause. For a question bank consumed as markdown, the practical questions are whether you are redistributing the files and whether you are offering a modified version as a network service. A personal clone for study raises neither. Republishing the quizzes inside a hosted practice site is where the network clause becomes relevant. The MIT declaration in package.json, if it were the governing licence, would allow far more permissive reuse, but package.json is not the licence file, and the two cannot both be correct.

This is not legal advice. If you plan to ship the content, read LICENSE yourself and, if the discrepancy matters to your use, get proper counsel. For individual study the question is largely academic.

When to reach for something else

If your goal is to demonstrate a skill on a profile, this repository cannot help, because the README states the assessments no longer exist on LinkedIn. A hands-on alternative such as building a small project in the language, or working through a general practice platform that generates fresh questions, gives you feedback that a static answer file cannot. The difference in approach is that those tools grade you against questions you have not seen, while this repository shows you the questions and the answers together, which is better for review and worse for testing yourself honestly.

If you need current, verified answers for a specific tool, a project's own documentation is the better source. The README's own contribution guidance points the same way: it asks contributors to attach an explanation or reference link to each answer so readers can check the reasoning instead of memorising a letter.

This repository is the right tool when you want breadth across many unrelated topics in one consistent format, offline, with translations. It is the wrong tool when you need a single authoritative answer, a graded result, or a credential.

Maintenance, releases and the cost of keeping it current

The last push to the default branch was on 2026-08-29, so the repository is still receiving changes. It is not archived. The release history is a different story: the only release listed is 1.0.4, tagged 2021-02-03. Content updates arrive as commits to main, not as versioned releases, so there is no changelog to read when you want to know what changed in the Git quiz between two points in time. You would have to read the commit history for that directory.

The package.json carries version 2.0.0, which does not correspond to any published release in the list. Treat the version fields as unrelated to the content's freshness.

Upgrade cost is low and unusual: you pull the repository again. There is no migration, no schema, no dependency tree beyond Prettier. The real cost is verification. Because the content is community-maintained across dozens of directories and the README's review column names specific people per topic, the effort of confirming that a given file is current falls on you. Running npm run format after a pull will keep your local copy consistent, but it will not tell you which answers survived a tool's latest release unchanged.

Editorial conclusion

Use this repository if you want a browsable, offline set of practice questions per topic, or if you want to contribute corrections and translations through the linked source repository. Do not use it as a substitute for a live LinkedIn assessment, because the README states those were discontinued in December 2023, and do not treat any answer file as authoritative: the table of contents itself marks entries such as Accounting as needing updates. Before relying on a topic, verify that the corresponding directory exists, that the quiz file lists a question count matching the answers column, and that the licence terms in LICENSE fit how you intend to redistribute the content.

Frequently asked questions

Official sources

1. Ebazhanov/linkedin-skill-assessments-quizzes on GitHub
2. License: AGPL-3.0
3. Project website
4. README
5. Releases