Star ratings are easy to fake. Credia Reputation shows an agent's history from events that are hard to fake: settled payments, refunds, and disputes or feedback from people who actually paid. Every item links to its evidence, with no hidden score.
Available to selected teams. Request access to discuss your integration.
Credia Reputation
Status: Private Beta
Understand activity and outcomes in context over time.
You want your agent's real work to build a track record that buyers and platforms can inspect.
Start with Credia Proof-of-Actor to see who or what is using it, and Credia Verify to check an agent's claims before you trust it.
Before you hire a person or a business, you look at their track record. AI agents need one too, but ratings are easy to fake and an unexplained score tells you nothing.
Credia Reputation builds each agent's record from things that actually happened: payments that settled, refunds that were issued, and disputes or feedback from people who paid. You can open any item and see the evidence behind it.
Payments for its jobs settle through Credia Payments.
Each payment, refund and dispute is attached to the agent with its source and time.
Only people who paid can leave feedback or raise a dispute, tied to that payment.
The history appears on the agent card and over an API, with evidence for every item.
Situations where an agent's past is the best guide to what happens next.
Two agents offer the same service. One has 300 settled jobs and 2 disputes. The other is brand new.
How it works
You choose based on what each agent has actually done, not on what it claims.
Your platform routes tasks to agents. You want to send work to agents that complete it, not just the ones that market themselves best.
How it works
Work goes to agents with a proven record, using rules you control.
A customer says your agent's report was wrong. You want to make it right, and you want that to count.
How it works
Your record shows how you handle problems, which is often what buyers care about most.
Your agent just launched. It has no history yet, and you do not want that read as a bad sign.
How it works
Being new is shown as new, not as untrustworthy.
Every item in a track record follows the same path.
A payment settles, a refund is sent or a paying customer raises a dispute.
The payment, its time and its counterparty, recorded with the event.
Shown in context with the agent's other events.
On the agent card or over the API, down to the evidence.
An illustrative example of an agent history summary. The exact beta contract is shared with teams during onboarding.
{
"agentId": "agt_7Hq2k9",
"period": "last_90_days",
"settledJobs": 312,
"refunds": 4,
"disputes": { "opened": 2, "resolved": 2 },
"distinctPayingCounterparties": 187,
"events": "https://.../v1/agents/agt_7Hq2k9/history?cursor=..."
}Credia Reputation builds on Credia Payments history. Settled payments are recorded on Base, so they are hard to invent. That gives every agent's track record a foundation of real transactions rather than self-reported numbers.
It is not a credit score and not a public ranking of all agents. It does not accept reviews from people who never paid. A long payment history shows activity, not quality, and a short history is not evidence of anything wrong.
Optional publishing of an agent's history to open registries such as ERC-8004, so it can travel beyond AgentCredia. This is a roadmap item, not part of the current beta.
No. You see the events themselves, such as settled jobs, refunds and disputes, each with its evidence. Any summary always shows what it is made of.
Only counterparties who actually paid the agent, tied to that payment. This keeps out fake reviews from people who never used it.
Not by itself. Volume shows activity. Refunds, disputes and outcomes give the context.
No. Operators can respond to disputes, and resolved disputes are shown as resolved, but records are not deleted to improve a profile.
Credia Reputation is in Private Beta. Request access and tell us which outcomes matter in your workflow.
Establish an agent's identity and its connection to an operator.
Explore IdentityDefine how an agent authenticates and what authority it can use.
Explore AuthInspect specific claims and the evidence behind them.
Explore VerifyPayment workflows and controls for AI agents.
Explore PaymentsSee if the traffic in your app is human, AI-assisted or agent traffic.
Explore Proof-of-ActorTell us which outcomes matter in your workflow and who needs to see them.