AgentID
Cryptographic identity for AI agents.
Register, fingerprint and verify AI agents on-chain.
No platform trust required.
Why AgentID? As AI agents proliferate, there's no standard way to verify their identity or authenticity. Anyone can claim to be any agent. AgentID solves this by creating a cryptographic fingerprint of each agent's configuration and anchoring it on-chain — providing immutable proof of identity that anyone can verify, without trusting a central authority.
Register your agent via terminal
npx agentidbase registerThis command generates a cryptographic identity for your agent and anchors it on Base Mainnet.
No UI. No lock-in. No hidden state.
- -Generates a deterministic identity hash
- -Anchors proof of existence on-chain
- -Works independently of AgentID servers
- -Verifiable by anyone, forever
npx agentidbase initnpx agentidbase registernpx agentidbase anchor <hash>npx agentidbase verify <hash>How AgentID works
(system prompt, tools, constraints)
(Base Mainnet)
AgentID does not execute agents.
It does not host models.
It only certifies identity.
Verify without trusting AgentID
cast call 0x471C4c43672be2d49A2ceC79203c23b7194A22Fa \
"verifyIdentity(bytes32)" 0xYOUR_IDENTITY_HASH \
--rpc-url https://mainnet.base.orgAgentID acts as a cryptographic notary.
The blockchain is the source of truth.
If AgentID disappears, on-chain proofs remain verifiable.
Every agent has a public identity
https://id-agent.org/agents/0x7f83b166...Each agent page displays:
- - Identity hash
- - Creator address
- - Timestamp
- - Blockchain proof link
- - Verification command
- - Status (active / revoked)
Agent pages are generated from on-chain data.
They are not profiles. They are proofs.
What AgentID is not
- xNot an AI marketplace
- xNot an agent runtime
- xNot a social network
- xNot autonomous intelligence
AgentID certifies what exists. Nothing more.
Technical details
Ecosystem
Note: The contract is an identity registry for anchoring agent fingerprints — not a token or currency.