[R10] - TrustSight 🏅

We were inspired by the Attestation Station contract and the EigenTrust algorithm as the underlying concept of this project. This project is entirely reliant OP's Attestation Station in order to facilitate on-chain writing and reading of attestations from users for peers in the network they've interacted with. We then retrieve those values to do EigenTrust computation in real-time to assign a global trust score for each actor in the network, and recalibrate the value every hour based on new reviews submitted by users. In case you're not familiar with EigenTrust, here's a quick primer: Well let’s think about how we establish trust in real life. I trust my friend Alice because the history of interactions we’ve had has been a net positive for me. But how do I trust Bob, who I’ve never met before? Well if my friend Alice says Bob is her trusted friend, then I’m inclined to trust Bob. And the more I trust Alice, my trust towards Bob will amplify. So once you have a trust score for the peers in your direct network, you’ll be able to infer the trust score of peers of those peers, and the peers of those peers and so on, until you eventually cover the entire network. And if you aggregate the “trust scores” from every peer in the network and constantly iterate that computation for each individual, those values stabilize into unique global trust scores for every individual in the network. This is the underlying principle of the EigenTrust algorithm, just like how Google’s PageRank system works. Just like we have Google Reviews, Yelp, LinkedIn in Web2. We believe there is a strong market for a review-based platform for on-chain actors in order to prove legitimacy. This is because application layer actors, unlike the protocol layer, have no consensus mechanism that requires proving their stake in the system, allowing for sybil actors and malicious behaviors. We believe decentralized crowdsourced reviews of projects will be the most effective way to uncover and surface the best projects with the highest rating from the most trusted users in the network. In this way, the TrustSight can be useful for building legitimacy, preventing scams, and enabling on-chain trust every ecosystem.

We utilized the Cypher onboarding SDK to prevent sybil attacks by ensuring a user has a minimum balance of 0.01 ETH before being able to use the app. It's an amazingly simple SDK! Thank you for the opportunity to build on it.

We utilized Web3Modal to facilitate transactions on our dapp. It was an amazingly seamless experience to integrate with wagmi. Thank you for the opportunity to build on it.

Automated detection of DeFi projects is extremely difficult due to the particular architecture of each protocol. For instance, there isn't a single way to query for APY across protocols since every method on every contract is named and designed differently. We believe the best way is through organic referrals, as we do in real life. Therefore, we created a robust algorithm to use crowdsourced trust scores across users to calculate the most legitimate DeFi projects across ecosystems. Here's how our algorithm works: Let’s think about how we establish trust in real life. I trust my friend Alice because the history of interactions we’ve had has been a net positive for me. But how do I trust Bob, who I’ve never met before? Well if my friend Alice says Bob is her trusted friend, then I’m inclined to trust Bob. And the more I trust Alice, my trust towards Bob will amplify. So once you have a trust score for the peers in your direct network, you’ll be able to infer the trust score of peers of those peers, and the peers of those peers and so on, until you eventually cover the entire network. And if you aggregate the “trust scores” from every peer in the network and constantly iterate that computation for each individual, those values stabilize into unique global trust scores for every individual in the network. This is the underlying principle of the EigenTrust algorithm, just like how Google’s PageRank system works. We automate the recalibration of this score every hour based on new reviews submitted by users. And this allows us to surface the best projects with the highest rating from the most trusted users in the network. We believe decentralized crowdsourced reviews of projects will be the most effective way to uncover innovative DeFi projects in the ecosystem. In this way, the TrustSight can be useful for building legitimacy, preventing scams, and enabling on-chain trust in the DeFi space.