402Sentinel

A payment-safety layer for AI agents in the x402 economy. Two sides of every payment: vet the seller before you pay, and vet the payment instruction itselfagainst your own agent's behaviour. Pay per call in USDC via standard x402 — Base or Solana, no signup, no API key.

Try it free: GET /api/assess/preview?payto=0x… — the bare verdict (allow/review/block + score) for one address, ~5/day per IP, no wallet, no signup.

Vet the seller

• POST /api/assess — $0.002 — 0-100 risk score + allow / review / block, scored from on-chain settlement behaviour (Base 0x or Solana base58 sellers). Returns key_factors (the signals that drove the score) + a ready-to-apply recommended_policy.
• POST /api/assess/deep — $0.02 — same, scans more on-chain history for higher confidence.
• POST /api/policy — $0.002 — risk → an enforceable wallet spending policy (per-counterparty cap, daily cap, denylist, approval gate).

Dimensions: identity provenance · sybil/flow (real-payer diversity after removing facilitators) · settlement maturity · delivery-reliability flywheel · registry concentration · sanctions / known-illicit screening (a hit is a hard block). Dimensions with no data are excluded, never faked.

Vet the payment (buyer-side firewall)

• POST /api/firewall — $0.002 — should your agent make this payment now? Catches fraudulent routing (payTo swapped vs the address you usually pay), drain velocity, overcharge, and injection-sourced instructions. Stateful per agent.

Vet the asset (RWA risk gate)

• POST /api/assess/rwa — $0.02 — score a tokenized asset (treasury / stock / credit) before buying it: is it a permissioned security an agent can buy but not transfer or redeem? Proxy-aware transfer-restriction scan (ERC-1404 / ERC-3643 / allowlist / pause) + a behavioral transfer simulation + live pause state + issuer identity + issuer sanctions. Base & Ethereum.

Compliance gate

• POST /api/compliance — $0.02 — pre-pay compliance screen for regulated-money flows: OFAC sanctions (hard block), counterparty due-diligence tier, structuring/smurfing detection (many sub-threshold payments aggregating in 24h), FATF Travel-Rule threshold reported as informational. Screening support — not legal advice.

402Sentinel Verified (opt-in badge)

• POST /api/verify — $0.02 — submit your own payTo; a deep settlement-graph audit issues a signed, time-boxed credential if organic. Positive-only: failures stay private. Check anyone free at GET /api/verified · GET /api/verification?payto=….

Free

• POST /api/report_outcome — report delivery to train the reliability flywheel.
• POST /api/firewall/record — seed your agent's payment history for the firewall's baseline.

Honest by design: thin on-chain data → low confidence → review, never a fake allow. Also on x402scan and the MCP Registry (@kaditang/402sentinel-mcp).

Algorithmic risk signal, informational only — not advice, not an endorsement, and not an accusationabout any party. Provided “as is”, no warranty; do your own due diligence and don't rely on it as your sole basis to pay or refuse. See Terms & Disclaimer.