The First AML Check an Agent Can Buy Should Be a Full Score

An agent in a deal room does not need a lecture about crypto compliance. It needs an answer before money moves.

The user can say: check this wallet on Sanctuary and buy the full AML review. The agent can request the paid check, pay the small crypto amount, and return the full score.

The free bot check remains the front door. It is useful for quick triage.

The agent-paid flow is for the full AML check: score, risk level, evidence categories, and decision context. That matters when the next step is a payout, trade, refund, allocation, or treasury move.

Most AML products assume a human sitting in a dashboard. Agents do not work that way.

They run inside tasks. They receive a wallet, a spend limit, a policy, and a goal. If screening requires a human checkout, the workflow breaks. If the agent can buy one full check, the workflow stays alive.

OpenClaw can use it. HermesAgent can use it. Codex can use it. Any runner that can call a tool and pay can use it.

Telegram trades, contractor payouts, OTC desks, token launches, escrow reviews, marketplace refunds: the pattern is the same. Wallet first. Score before value moves.

To our knowledge, Sanctuary is the first AML service built so an autonomous agent can buy a full wallet check directly.

That is the product shift. AML screening stops being a dashboard action only. It becomes a paid step inside agent work.