Back to Projects
React
Python
FastAPI
Javascript
Tailwind
ML

Kraven The Hunter

An ML powered chrome extension that parses the url and sends it to the an ML endine and database to check if the it is malicious or not.

Kraven The Hunter

Project Overview

So Tyrone, why did you decide to do this project? I plead the 5th. What I will say, is that the idea to do it was pretty cool, am I right? LOL. This is essentially a browser extension that parses the URL the browser is currently on and sends it somehwere..

Send it where? Well, I have a FastAPI server on the backend exposing a machine learning model trained on a databases of known malicious URL. First I had to decide what are my features. Well, for starters, malicious URLs tend to be werid, right? What do I mean by that? They are long and have lots of special characters. Well, So I used the length of the URL and the number of special characters and as well as a host of other stuff you know.

On the other hand, if we have a filter feature enabled on the server, we can direct traffic to check whether the URL is a shopping site or an adult entertainment site and all. Why would someone want to filter those out? Psshhhh, I don't know, maybe you have kids and you don't want them to see that stuff. So I added a filter for that too.

The page we eventually direct the user to is built off of React and Tailwind and it shows a minimal report on whether the URL is malicious or not and our degree of accuracy. If the URL is good, no redirects.

Key Features

  • Auto URL parsing.
  • Typo-squatting and phishing detection.
  • Shopping cart and wishlist functionality
  • Secure payment processing with Stripe
  • Order management and tracking
  • Responsive design for all devices

Challenges & Solutions

Implementing ML without using an established ML library like tensorflow.

Capturing the URL and redirecting to a warning page if malicious. Chrome doesn't like that.

Creating a responsive design that works across all device sizes

Setting up automated deployment pipelines

Key Learnings

Definitely how to build a browser extension and browser APIs.
Not to forget how to develop ML models without using an established ML library.
To be honest, I already knew how to code pretty well with everything else.

Project Info

Timeline

3 months

Team

Solo project

Status

Completed(But Imperfect)

Tech Stack

Frontend

React
TypeScript
Tailwind CSS
Javascript

Backend

Python
FastAPI
JWT Authentication

Database

N/A

Payment

N/A

Deployment

N/A

Tools

Git
GitHub Actions
Postman
VS Code