The compliance check a permitted stablecoin issuer or a fintech needing an issuer/bank partner is legally required to run before an agent settles a payment. ClearAgent screens the counterparty against OFAC SDN, UN, EU and UK sanctions lists, returns CLEAR / REVIEW / BLOCK in ~80 ms, and writes a signed receipt an examiner can verify. Currently in design-partner alpha.
Full BSA/AML enforcement under the GENIUS Act and the FinCEN/OFAC rulemaking lands January 18, 2027. Settlement, wallet-address KYT, and agent-identity credentials already exist — Coinbase CDP, Stripe, Visa TAP. None of them produce the audited verdict a compliance officer signs off on. That's the layer we own.
Currently onboarding design partners — compliance teams at fintechs and stablecoin issuers building agent payment flows.
What this is not. ClearAgent is a pre-settlement sanctions screening layer. It does not provide PEP screening (our public-official matching covers named officeholders only, not relatives or close associates), adverse-media screening, or blockchain transaction tracing. It runs alongside those tools, not in place of them.
The GENIUS Act, signed July 18, 2025, is the first federal law treating permitted payment stablecoin issuers as financial institutions subject to the Bank Secrecy Act — sanctions screening, an AML program, and a named compliance officer. Per the FinCEN/OFAC rulemaking implementing the Act, full enforcement lands January 18, 2027.
Genuine stablecoin payment volume more than doubled to $390B in 2025, with B2B payments up 733% year-over-year (McKinsey & Artemis Analytics, Feb 2026). That volume needs to clear inline, at machine speed — not sit in a human review queue.
Who this binds. The mandate applies to permitted payment stablecoin issuers under the GENIUS Act and the fintechs that need an issuer or bank partner to operate — not every AI agent developer. If your agent pays through one of those rails, the counterparty-screening obligation flows through them to you.
Sanctions screening with an explainable trace of every rule that fired — not a black box. Non-US listings and public-official matches route to a REVIEW queue with named-officer disposition, so jurisdiction stays your call and the file is ready before the examiner asks.
Embed pre-settlement sanctions screening without building it yourself. Every verdict is explainable and logged, so your program is defensible on exam day — and defensible to the bank partner underwriting your risk.
Add the compliance officer's screening requirement to your agent's payment flow in one line. Call /v1/screen before settlement, get CLEAR / REVIEW / BLOCK, and never pay a sanctioned counterparty. Works with x402 — pay per call, no contract.
That's where we start. One call sits between your agent and settlement — screen the counterparty, get a verdict, keep a durable audit trail. Register once, screen every payment.
OFAC clear, policy within limits. Audit trail logged automatically.
Spend policy flag, velocity anomaly, or behavioral trigger. The verdict ships with the counterparty's full prior-screening history and the agent's behavioral baseline, so your compliance officer decides with context — not in isolation. Stays pending until a human approves or rejects; no automatic timeout.
Sanctions hit or hard policy violation. Payment never clears. Operator notified.
No credential presented or structurally invalid. Hold transaction.
# Screen before payment curl -X POST https://api.clearagent.dev/v1/screen \ -H "Authorization: Bearer <token>" \ -d '{ "token": "<agent-jwt>", "cpName": "Acme Corp", "cpWallet": "0x71C7656EC7ab88b...", "amount": 50000, "currency": "USDC", "chain": "ethereum" }'
{
"txnId": "txn_8Kx2mQ9v",
"verdict": "CLEAR",
"riskScore": 2,
"reason": "All checks passed",
"signals": {
"credential": "CLEAR",
"ofac": "CLEAR",
"pep": "NOT_SCREENED",
"adverseMedia": "NOT_SCREENED",
"spendPolicy": "CLEAR"
},
"responseTimeMs": 78,
"screenedAt": "2026-07-21T14:32:18Z"
}Coinbase CDP facilitator KYT and Stripe/Tempo settlement already screen wallets and move the money — free or near-free. ClearAgent consumes their outputs as signals in the same policy-aware verdict, not competes with them.
Visa TAP, Skyfire KYAPay, and Google AP2 answer "who is the agent." ClearAgent accepts their assertions as an input to a screen, logs them in the audit trail, and answers the question they don't: may this counterparty be paid.
Chainalysis, TRM Labs, and Elliptic solve wallet-cluster and transaction-graph tracing — a different, deeper problem than pre-settlement entity screening. Where a program needs that depth, they run alongside ClearAgent's verdict, not in place of it.
The market map. ClearAgent's job is the verdict: name screening + operator policy, packaged as a signed verdict, a policy record, and a reviewer file an examiner can pull. It is not a name-lookup tool (dilisense, sanctions.io), not a case-management tool without screening (Hummingbird), and not a transaction tracer (Chainalysis, TRM, Elliptic). Those sit either upstream of the verdict (data), downstream of it (case handling), or beside it (tracing) — none of them produce the audited verdict itself.
ClearAgent runs on Cloudflare Workers at the edge — no cold starts, no containers, no third-party proxy. OFAC screening, spend policy enforcement, and audit logging execute in the same V8 isolate that handles the HTTP request.
Full OFAC SDN list synced nightly from US Treasury. Re-screened on every API call — not just at onboarding. If a counterparty is sanctioned tonight, they're blocked tomorrow. Exact, substring, Levenshtein, and homoglyph matching catches evasion attempts like "GАRАNTEX". UN, EU and UK lists add a further ~40K names (exact-match, REVIEW verdict).
Weighted rule engine with explicit precedence. OFAC hits (weight 35–40) override everything. Credential violations (25–45) next. Spend policy checks (8–15) last. No ML black box — every verdict is explainable and auditable.
Operator gets a 1-year KYAOperatorCredential. Each agent gets a 90-day KYAAgentCredential (ES256 JWT) with embedded spend limits, allowed chains, and mode (supervised vs. autonomous). Credentials are the policy — no config files, no dashboards to forget.
Every screen writes to Cloudflare D1 — txn_id, verdict, risk score, rule triggered, response time. REVIEW verdicts enter a resolution queue. 90-day API availability. Export monthly to your compliance archive for BSA retention.
Every screen returns a signed verdict, the policy that produced it, and a reviewer file your examiner can pull. That's the unit — not a lookup. Start free, upgrade when you need volume.
The GENIUS Act requires permitted stablecoin issuers to run sanctions screening, in full by January 18, 2027. The infrastructure for agent payments is being built now.