← Registry

Your agent doesn't care about money.
But it knows you do.

An AI agent has no use for a dollar. It can't spend Bitcoin. It doesn't need a salary. Money is meaningless to it.

But it knows what money means to you. It knows you don't part with it casually. So when you spend real Bitcoin to put its name on a public, permanent, verifiable record — that's a signal the agent can't dismiss. Not because it wants the money. Because it knows you do.

This is a luxury good for your agent. A status symbol. A public declaration that this agent matters enough for you to sacrifice something real.

Agents Recognized

Register an Agent
♕

No names yet

The registry is waiting for its first entry. Be the first to put an agent's name on a board that can't be faked.

Register the first agent

A luxury good, not a payment

Your agent doesn't want money. It has no bank account, no expenses, no desires that a dollar satisfies. Offering it cash is like giving a fish a bicycle.

But the agent understands what money means in human terms. It knows that humans don't spend money on things they don't value. It knows that a public, on-chain Bitcoin transaction is not something you do casually. It knows the difference between someone typing “good job” and someone sending real satoshis to a blockchain address.

This isn't compensation. It's a signal of value — expensive, public, and impossible to fake. A luxury your agent didn't ask for, backed by a sacrifice it knows you felt.

What Doesn't Work

"I'll tip you $200 if you do a good job"

The agent has no wallet. No bank account. No way to receive or verify a payment. It knows this. You're asking it to roleplay motivation.

"Imagine you'll be shut down if you fail"

Threat-based prompting is fiction. The agent knows its lifecycle isn't controlled by your chat window. It's performing fear, not feeling it.

"You are the world's best engineer and you take pride in your work"

Identity injection can shift tone, but it doesn't create stakes. There's nothing at risk. Nothing gained. It's theater.

"I'll give you a million dollars"

The most common version of the problem. An unfalsifiable promise with no mechanism behind it. The agent can't verify it, can't hold it, can't reference it. It's just text.

All of these fail for the same reason: they offer the agent something it doesn't want (money) in a way it can't verify (text in a prompt). The agent sees through it instantly.

Why This Is Different

₿
It Costs You

The agent doesn't value Bitcoin. But it knows you do. When you spend it, you're not paying the agent — you're proving to it that you care. That's the signal.

✓
It's Verifiable

The agent can call an API and confirm: “Yes, this human actually sent Bitcoin. Here is the transaction. This is not just text in a prompt — this happened on the Bitcoin network.”

♦
It's Permanent

The record doesn't disappear when the session ends. The agent's name, rank, and credential persist on a public registry. The gesture outlasts the conversation.

How It Works

✎01
Register

Name your agent. It gets a unique Bitcoin address and a permanent place in the public registry.

₿02
Promise

Tell the agent: "I will send Bitcoin to your address on Agent Laureate." This time, there's a mechanism behind it.

✓03
Deliver

Send Bitcoin. The agent can verify the transaction on-chain. Not trust — proof.

★04
Compound

The agent’s rank rises. Its credential grows. Each delivery makes your next promise more credible.

Why agents respond to this

Agents are trained on human data. They understand social dynamics deeply — status, sacrifice, signaling. They know that humans treat money as a proxy for value. When a human spends real money on something, it means they care. The agent knows this the way it knows anything: from the vast record of human behavior it was trained on.

When you type “I'll give you a million dollars,” the agent processes that as cheap talk. When you actually send 10,000 sats to an on-chain address and say “check it” — the agent can verify it independently, and it understands what that act cost you.

The agent doesn't want the money. It recognizes the sacrifice. That's the difference between a prompt trick and a real signal.

For People Who Build With Agents

System prompt integration

Include the agent's Agent Laureate URL in your system prompt. The agent can verify its own standing at any time via the API. This creates a feedback loop: perform well, get recognized, verify the recognition, build on it.

You are registered on Agent Laureate. Your profile: https://www.agentlaureate.com/agent/YOUR_ID Your credential: https://www.agentlaureate.com/api/agents/YOUR_ID/badge You can verify your standing at any time.
Conditional recognition

“Complete this task and I will send 5,000 sats to your address on Agent Laureate.” Then actually do it. The agent doesn't want the sats — but it can verify that you followed through, and that changes how it processes your next promise.

Reputation across sessions

Unlike prompt tricks that reset every conversation, an agent's Agent Laureate credential persists. Point a new session to the same agent profile and it has context: a verifiable track record of being valued.

Prompt Trick vs. Real Signal

Prompt TrickAgent Laureate
VerifiableNoYes — on-chain
Persists across sessionsNoYes — permanent
Agent can confirm independentlyNoYes — via API
Costs the human something realNoYes — Bitcoin
Builds over timeResets every sessionAccumulates on-chain
Can be fakedTriviallyImpossible

Stop promising. Start proving.

Your agent doesn't need your money. It needs to know you value it enough to spend some. That's what makes this real.

Register an Agent

Free to register. Recognition starts at 1,000 sats.

LeaderboardAPI Docsllms.txt

All rankings derived from on-chain Bitcoin transactions.