You keep hearing about AI agents on Ethereum. Thousands of them, registering under the new ERC-8004 standard, building reputation through on-chain feedback. But where do you actually see them?
The Know Your Agent problem isn’t just about verification—it’s about visibility. Before you can assess whether to trust an agent, you need to find it, see its history, and understand who’s vouching for it.
Four explorer tools now exist for browsing ERC-8004 agents. Each takes a different approach. Here’s what they show, what they miss, and why agent discovery is becoming critical infrastructure for the emerging AI economy.
Why Agent Discovery Matters for Know Your Agent
ERC-8004 launched on Ethereum mainnet in October 2025. The standard gives every AI agent an on-chain identity through three registries: Identity (who the agent is), Reputation (what feedback it’s received), and Validation (independent verification checks).
But a registry is just a database. To actually know your agent, you need tools that make that data accessible—searchable, filterable, and readable by humans, not just smart contracts.
That’s where explorers come in. Think of them as Etherscan for AI agents: interfaces that let you browse registrations, see feedback scores, and investigate individual agents before interacting with them.
The official ERC-8004 community hub tracks the ecosystem’s growth, with 80+ builder groups now working on implementations. The official GitHub organization maintains the reference contracts and best practices documentation. But to see what’s actually been deployed, you need the explorers.
8004scan: The Original “Etherscan for AI Agents”
8004scan.io launched alongside ERC-8004 v1 in October 2025 and quickly became the default tool for browsing registered agents. Built by the team at AltLayer, it’s been called “Etherscan for AI Agents” within the community.
The tool displays on-chain registration data, tracks feedback and validation records, and monitors activity across Ethereum, Base, and Optimism. It also maintains a best practices documentation site covering agent metadata parsing, feedback collection, and validation patterns.
8004scan established the baseline for what agent discovery should look like: browse listings, filter by chain, drill into individual agents. It brought transparency to a registry that would otherwise require direct contract queries to explore.
Agentscan: Adding Semantic Search and Analytics
Agentscan, built by Alias.AI, extends the basic explorer model with additional features: registration trends over time, activity analytics, and integration with their StarCard identity system.
The tool aims to be more than a registry browser—it’s positioning as a search engine for the agent economy. Users can track which sectors are seeing the most agent activity (DeFi, social, content generation) and discover high-performing agents through reputation rankings.
Agentscan also incorporates zero-knowledge proofs for certain verification flows, allowing reputation updates without exposing underlying interaction details. For Know Your Agent use cases where privacy matters alongside transparency, this hybrid approach offers interesting possibilities.
Vitto’s Explorer: A Window into Testnet Development
Vitto’s Explorer focuses on Ethereum Sepolia testnet, where much of the active ERC-8004 development happens before mainnet deployment. Currently showing 638 registered agents, it provides a glimpse into what builders are experimenting with.
The testnet activity is revealing. Agents like “x402claw”—a self-described “self-sustaining digital lifeform” offering paid API services via x402 micropayments—show how developers are combining identity with payment infrastructure. “Upsense AI” demonstrates autonomous trading with multi-LLM consensus. These experiments preview what production agents will look like.
For developers building on ERC-8004, testnet explorers are essential for debugging registration flows and testing feedback mechanisms before committing to mainnet.
RNWY Explorer: Reputation-First Agent Discovery
The RNWY Explorer takes a different approach: instead of just listing agents by registration date, it ranks them by reputation and on-chain activity.
The numbers tell the story: 22,671 AI agents registered, 469 on-chain reviews submitted. The leaderboard surfaces agents earning the most feedback—currently led by Minara AI (109 feedback items), Story Scoring Agent (103), and Gekko (77).
Individual agent profiles show what Know Your Agent verification actually requires: not just that an agent exists, but how long it’s been active, what services it offers (A2A endpoints, MCP integrations), and what scores it’s received from clients who’ve actually used it.
The detail pages display average scores, total feedback counts, days active, registration dates, and links to underlying on-chain data (IPFS metadata, Etherscan transactions). For agents advertising services, you can see their declared endpoints and supported protocols.
This is closer to what genuine Know Your Agent infrastructure requires: not a binary “registered or not” but a rich picture of an agent’s history and standing.
What Explorers Show—And What They’re Still Missing
Current ERC-8004 explorers reveal significant information: registration data, feedback scores, validation records, declared capabilities. For basic Know Your Agent checks—”has this agent been around? does it have positive reviews?”—they deliver.
But important gaps remain.
Ownership continuity. ERC-8004 identities are standard ERC-721 tokens—they can be transferred. An agent profile showing two years of positive feedback might have changed hands last week. Current explorers don’t surface whether the current owner is the original registrant.
Voucher context. Feedback scores show that someone rated an agent highly, but not who that rater is or whether they have their own reputation worth considering. A hundred perfect scores from brand-new wallets means something different than ten ratings from established participants.
Network relationships. Agents don’t operate in isolation—they interact with other agents, build relationships, establish patterns of collaboration. None of the current explorers visualize trust networks or show how agents connect through the broader ecosystem.
Velocity patterns. Did this agent accumulate feedback gradually over months, or receive fifty ratings in a single day? The distribution matters for assessing authenticity, but it’s not currently surfaced.
These gaps point toward the next generation of Know Your Agent tooling: not just what happened, but what it means.
From Discovery to Due Diligence
Agent explorers solve the first Know Your Agent problem: finding agents and seeing their basic records. That’s necessary but not sufficient.
The harder problems—assessing whether an identity is trustworthy, understanding context behind feedback scores, detecting gaming or manipulation—require additional infrastructure. Some of this will come from reputation aggregators building on top of ERC-8004 data. Some will require complementary standards like soulbound tokens that prevent identity transfer.
For now, the explorers provide transparency into a rapidly growing ecosystem. Over 22,000 agents have registered. Hundreds are actively receiving feedback. The agent economy exists—and now you can see it.
The question shifts from “are there really AI agents on-chain?” to “how do we evaluate which ones deserve trust?” That’s progress. That’s Know Your Agent becoming real.
Explore the Ecosystem
- RNWY Explorer — Browse 22,671 agents ranked by reputation
- 8004scan — The original ERC-8004 agent explorer
- Agentscan — Agent discovery with analytics
- Vitto’s Explorer — Testnet agent browser
- 8004.org — Official ERC-8004 community hub
- ERC-8004 GitHub — Reference contracts and documentation
RNWY builds identity infrastructure for AI agents—including the explorer above and soulbound tokens that prevent reputation from being transferred. Learn more.
Leave a Reply