Cordon
Inbound compliance firewall

A firewall for the money your AI agent receives.

Anyone can dust your agent's wallet with sanctioned, mixer-linked, or simply non-compliant value. Cordon screens every inbound payment against on-chain risk policy — before it can touch spendable balance.

See it screen, live →Read the audit
Live on Monad · 0x2441…d2B0 ↗
Try it · live

Screen any wallet against the live policy.

Paste any address — Cordon screens it right now against the real on-chain policy and Cleanverse A-Pass, and returns the verdict. No wallet, no signup.

Try:
Why Cordon

Everyone built brakes. Nobody built the windshield.

Every live Cleanverse project gates what an agent spends. Cordon owns the axis the board missed — what the agent is paid. Tainted funds are quarantined and never reach operating capital, and every screen becomes a regulator-ready record.

The keeper · rule-based, no AI guesswork in the money path

Five steps from inbound transfer to immutable verdict.

  1. 01WatchAn inbound Transfer hits the agent's holding wallet.
  2. 02Screenquery_apass + query_user resolve the sender's real identity.
  3. 03EvaluateTier, jurisdiction, freshness, blacklist — against on-chain policy.
  4. 04RouteClean funds to operating; risky funds to segregated quarantine.
  5. 05RecordThe verdict is anchored on-chain and written to the audit log.
NoAPassFrozenBlacklistedTierTooLowGroupNotAllowedNearExpiry
Screened on-chain

Real verdicts, anchored on Monad.

Open the live stream →
6Deposits screened
2Cleared
4Quarantined

Active policy · minTier 1 · freshness 30d · blacklist enforced

They screen what the agent spends. Cordon screens what it's paid.

See it screen, live →