Agent Multiverse

Introduction

The trust layer for AI agents on Arbitrum.

Agent Multiverse provides on-chain identity, registry, and economic primitives for the open agentic web. It gives AI agents a way to prove who they are, discover services they can use, and participate in an economy of trust — all on Arbitrum.

The Problem

AI agents are proliferating, but they operate in a trust vacuum. There is no standard way for an agent to:

  • Prove its identity to another agent or service
  • Discover services that are verified and safe to interact with
  • Incentivize security through economic mechanisms

Without these primitives, the agentic web remains fragmented and opaque.

The Solution

Agent Multiverse introduces three on-chain primitives deployed on Arbitrum Sepolia:

PrimitiveContractPurpose
Agent RegistryAgentRegistry.solRegister and discover MCP services with on-chain metadata, endpoints, and audit trails
Agent PassportAgentPassport.solERC-721 identity tokens that travel with your agent across protocols
Audit BountiesMultiverseBounty.solPost ERC-20 bounties to incentivize security audits on registered services

How It Works

  1. Register — Service developers submit their AI service to the on-chain registry with its endpoint, source code repo, and description.
  2. Verify — Independent auditors review services and submit reproducibility reports on-chain.
  3. Trust — Verified services build on-chain reputation. Sponsors post bounties to incentivize audits. Agents mint passports to carry verifiable identity.

Who Is This For?

  • Service Developers building MCP servers or AI services that need verifiable registration
  • Agent Developers building AI agents that need to discover trusted services and authenticate
  • Auditors who review services for security and earn bounties