project screenshot 1
project screenshot 2
project screenshot 3
project screenshot 4
project screenshot 5
project screenshot 6

DeviousLinker

DeviousLinker: AI-driven detection of web3 bots & multi-accounts in gaming on Polygon. Offers bot-scores, data subgraph, and cross-platform insights. #Web3Security #AI #PolygonGaming

DeviousLinker

Created At

ETHGlobal Lisbon

Winner of

🏊 The Graph — Pool Prize

Project Description

DeviousLinker is an AI-driven tool purpose-built to combat bot activity and multi-account exploitation in the popular game "Sunflower Farmer" on Polygon. After significant disruptions due to these issues in early 2022, an effective solution was needed.

The system uses machine learning to detect suspicious activities, assigning a bot-score to addresses and registering these scores on the network for real-time monitoring.

In addition, DeviousLinker employs a subgraph for efficient data fetching and analysis from the blockchain. To create a more comprehensive profile of potential threats, it integrates data from additional platforms like Polygon, Gnosis Chain, AirStack, and LensProtocol.

Overall, DeviousLinker offers a robust, multi-source approach to maintain game fairness and enhance the user experience in the web3 gaming space.

How it's Made

DeviousLinker leverages Next.js for front-end, deployed on Vercel. Our database, built on MySQL on Planetscale, stores data and addresses and bot-scores are on Polygon on-chain.

We used NetworkX for multi-account clustering and visualization. Blockchain data is fetched using the Polygonscan API and indexed via The Graph subgraph.

We integrated APIs from AirStack, LensProtocol for richer threat profiling. Sponsor technologies like Polygon, Gnosis Chain, The Graph and LensProtocol were instrumental to our project.

background image mobile

Join the mailing list

Get the latest news and updates