Cryptographic Integrity & Security Architecture
OwnPay operates a zero-custody, intent-driven settlement architecture on Base (Chain ID 8453). We enforce strict cryptographic invariants across human checkouts, retail POS terminals, and autonomous AI agent operations.
100% Non-Custodial Architecture
OwnPay never takes custody of merchant private keys or settlement funds. Payments settle directly from customer wallets or agent smart accounts into your merchant vault address on Base.
15-Point Onchain Verification
Every payment intent undergoes mathematical confirmation: transaction hash validation, Base block inclusion, sender verification, token contract address check (Circle Native USDC), and slippage ceilings.
Double-Entry Settlement Ledger
Every transaction, fee deduction, and net allocation is recorded into an append-only cryptographic ledger where Total Debits strictly equal Total Credits. No hidden fees, no balance drift.
AI Agent Policy Guardrails & Kill-Switch
Autonomous AI agents are bound by 3-tier risk policies: per-transaction caps, daily spend limits, and recipient whitelist contracts. Merchants can trigger an instant 1-click key freeze via API or dashboard.
HMAC-SHA256 Webhook Signatures
All outgoing webhook notifications are cryptographically signed with unique merchant HMAC secrets and timestamp headers to eliminate man-in-the-middle attacks and replay attempts.
Paymaster Rate Limits (EIP-4337)
Account abstraction sponsorship through Alchemy Gas Manager incorporates strict velocity caps, preventing drain attacks and enforcing daily gas sponsorship budgets per merchant profile.
Responsible Vulnerability Disclosure
We take security seriously and appreciate ethical security researchers who discover vulnerabilities and report them responsibly. If you believe you have found a security vulnerability in OwnPay's core engine, API endpoints, or smart contract interactions, please report it immediately:
- Provide detailed reproduction steps or proof-of-concept scripts.
- Do not attempt to access private merchant accounts, sensitive customer data, or drain funds on Base Mainnet.
- Allow us adequate time to investigate and patch before making any public disclosure.
- Our security engineering team acknowledges legitimate reports within 24 business hours.