npmsmell is a project I started to help reduce the risk of supply chain attacks in the NPM ecosystem. Many widely used packages depend on outdated or unnecessary dependencies, making them a prime target for supply chain attacks.

The site is all about education. In addition to metadata like the last update date and total transitive dependencies, it highlights viable upgrade paths. More often than not, the functionality these packages provide is now natively supported, thus hopefully lowering their download count and in turn making them less appealing for attackers.

The site is built with Astro and tailwindcss. The data is provided by my own Node Package Analyzer project.

Node Package Analyzer or npa gives you insight about a Node.js package.

• Number of dependencies
• Oldest dependency
• Licenses
• etc.

It is an open source Node.js application written in TypeScript.
Unit testing is done via Jest, code coverage is generated by istanbul
SonarQube is used to statically analyse the integrity of the code
Continuous Integration (CI) is provided by Azure pipelines
Every commit automatically triggers tests, code coverage and static code quality analysis.

During the development of the packageanalyzer I tried to validate its usefulness by building a little website where you could guess the total amount of dependencies for a particular package.
All the data for this project was gathered by the packageanalyzer
A goal of the project was to build a React application solely with function components to figure out their pros and cons compared to class components.

Special care was taken to provide A11Y.
Styling was done via CSS in JS framwork Emotion.
Unit testing is done via Jest.
E2E tests are handled via cypress.
Percy was used to provide visual regression testing.
Netlify was used to host the final application
CI workflow was setup with Azure and GitHub Actions.

In an effort to grow my vocabulary in Spanish I made a news site that would show me the news in Spanish, where I can easily translate a word and save it to my list of vocabularies

The application fetches RSS feeds from various news sites and provides translations via Google Translate

For the backend I used C# in conjunction with ASP.NET core to provide the API and authorization.
EF Core is used to supply the API with data.
The frontend is powered by React and React router written with TypeScript. MobX is used for state management.

Backend code is unit tested with xUnit.
Frontend integration tests are done via cypress.