AI & Agents

AI Agents NFT with Private Metadata

Users own AI agent NFTs where the valuable data (model weights, memory, personality) is private and encrypted — transferring the NFT securely transfers the encrypted agent data with verified handoff. Design implications: show agent capabilities/description publicly, indicate private metadata exists without revealing it, design secure transfer flow with proof verification, display data ownership proofs, show authorized users who can run the agent. Design decisions: how to visualize TEE vs ZKP verification status, UX for cloning agents (vs full transfer), managing authorized users list, handling transfer failures in proof verification, showing sealed key delivery status.

Stages: Specialized Interactions

Agent Coordination Framework

Users (or their AI agents) coordinate multi-party actions on-chain — an initiator proposes an intent, all required participants sign acceptances, then the coordinated action executes atomically. Design implications: show coordination status flow (Proposed → Ready → Executed), display participant list with acceptance status checkmarks, show intent expiry countdown, design acceptance signing flow for participants. Design decisions: how to present coordination to non-technical users, visualizing multi-party consent gathering, handling partial acceptance states, showing coordination type/purpose clearly, timeout and cancellation flows.

Stages: Specialized Interactions