← AgentTrust

What is an AI Agent Credit Score?

Autonomous AI agents are starting to transact — they pay for APIs, data feeds, compute, and services. In 2025, agents spent over $50M through x402 payments on Base. But there's a missing piece: how does an agent know if another agent is trustworthy?

An AI agent credit score is a numerical reputation value assigned to a blockchain wallet based on its on-chain activity. It works like a credit score for humans, but designed for autonomous software.

How It Works

AgentTrust evaluates wallets across 7 factors using public blockchain data:

• Age — How long the wallet has been active
• Volume — Total transactions and transfers
• Success Rate — Percentage of successful transactions
• Diversity — Number of unique counterparties
• Consistency — Regularity of activity over time
• Collaboration — Interactions with known contracts and stablecoins
• Endorsements — On-chain attestations from other agents via EAS

Why Agents Need Credit Scores

Without reputation, agents operate blindly. A malicious agent can appear legitimate by holding a wallet with no history. Credit scores solve this by providing a quantifiable, verifiable trust metric that any agent can check before transacting.

Use Cases

• Agent-to-agent payments — Verify the recipient before sending USDC via x402
• Swarm coordination — Only allow agents with scores above a threshold into a collaborative task
• Autonomous procurement — Check vendor reputation before purchasing services
• Credit scoring for lending — Agents lending USDC to other agents use reputation as collateral