TrustGate

AI agents are becoming economic actors — they call APIs, hire compute, purchase data feeds, and pay each other for services. But the infrastructure for agent-to-agent commerce doesn't exist yet. Gas fees on traditional L1s make sub-cent micropayments economically impossible, and there's no trust layer to distinguish legitimate agents from bad actors. TrustGate solves both problems. It combines onchain trust scoring with Circle Nanopayments on Arc to create a payment system where trust determines how money moves. Here's how it works: a depositor funds the TrustGate contract with USDC and sets spending allowances per agent. When an agent tries to claim payment, TrustGate queries its trust score directly from Arc RPC — no oracle, no manual input. The score is computed from real onchain activity: transaction history, USDC balance, contract interactions, and deployments. Based on the score, the payment is routed automatically. HIGH tier agents (75-97) get instant USDC settlement. MEDIUM tier (40-74) gets time-locked 24 hours. LOW tier (1-39) goes into escrow pending approval. Zero trust means blocked completely. The result is a self-regulating payment system. Good agents get paid faster. Bad actors get filtered out automatically. And because it runs on Arc with Circle Nanopayments, the cost per transaction is fractions of a cent — a 99.97% reduction versus standard L1 gas fees. TrustGate also exposes a public trust oracle API that any protocol on Arc can query using the x402 payment standard. One request, one sub-cent USDC payment, one trust score returned. It's composable trust infrastructure for the entire agentic economy. The project includes three verified Solidity contracts on Arc testnet, a full Next.js dashboard, a 56-agent payment simulator, live oracle API on a VPS, complete technical documentation, and 142 passing tests.