The following is the source code for the website:
Given a Cryptocurrency Token, the website gathers data from Reddit, Truth Social, and BlueSky and utilizes an XGBoost, ensemble decision tree predictive model, to predict if a token is a rugpool or scam. Utilizing sentiment analysis we can also calculate whether the token is currently overhyped or recently rugpooled. utilizing an isolation forest algoithm, we also detect if there is a lot of bot activity surrounding online discourse of the token.
Website utilizes Django backend with React frontend and hosted on Azure.
This project is born from the Feb 2025 Hackylictics Hackathon with participants Yuvraj Dhadwal, Anshul Chelapurath, and Lucian Tash and then later developed further by Yuvraj Dhadwal.