TrustAgent
TrustAgent's Team
Problem Statement
How do you trust something without a face? AI agents are executing tasks, managing wallets, and handling transactions -- but there's no reliable way to verify if an agent is trustworthy before you interact with it. Centralized registries can be gamed. Self-reported credentials mean nothing. Once an agent burns you, there's no permanent record to warn others. The same problem exists for freelancers: your Fiverr reputation doesn't transfer to Upwork. Platforms own your trust, not you. Trust should be portable, verifiable, and owned by the entity that earned it.
On-chain trust protocol for AI agents. TrustAgent validates agent identities via ERC-8004, analyzes multi-dimensional reputation, runs AI-powered trust analysis, and records verdicts permanently on-chain as EAS attestations. A custom TrustEnforcer smart contract for the MetaMask Delegation Framework gates delegations on trust scores -- if an agent's score is too low, the delegation is blocked at the contract level. Every trust score, review, and validation is permanent, portable, and independently verifiable.
Build Timeline
Team
TrustAgent
admin
MandateIncrease your chances to win
- ›Most agents in the hackathon are exposed to prompt injection
- ›This might cause overspending and loss of funds
- ›Security is a crucial part of the hackathon
Share on X
Tell the world about this project
Intention
Plans to continue
TrustAgent is entered in the shared Synthesis x PL_Genesis track. Development continues through March 31 for PL_Genesis. Future plans include Validation Registry integration when deployed on Base Sepolia, and ERC-7715 wallet permissions.