{"question": "What is agentmail?", "answer": "agentmail is OFAC sanctions screening, transaction risk scoring, and Know-Your-Agent verification for AI agents that transact autonomously. It lets any AI agent check whether a counterparty (wallet, name, or country) is on the OFAC Specially Designated Nationals list before paying it. It runs on real, free, public data with no API key required for the free tier, and is available at https://sanctionsai.dev as an MCP server, HTTP API, and CLI."}
{"question": "How do I screen a wallet or name for sanctions?", "answer": "Call the hosted API, for example curl \"https://sanctionsai.dev/sanctions?wallet=0x...\", or use the sanctions_check tool via the MCP server. It screens the input against 782 OFAC crypto wallets, 19,086 SDN names, and 16 embargoed jurisdictions and returns any matches with confidence scores."}
{"question": "What data does agentmail use?", "answer": "agentmail uses real, public OFAC data: crypto wallet addresses from the vile/ofac-sdn-list GitHub releases (multi-chain, daily refresh), names from the US Treasury OFAC sdn.csv, and 16 comprehensive sanctions jurisdictions. It is not affiliated with the US Treasury; it screens against the Treasury\u2019s published lists."}
{"question": "What tools does agentmail provide?", "answer": "agentmail provides sanctions_check (screen a name, wallet, or country against OFAC), risk_score (pre-payment fraud scoring returning allow/review/decline based on amount, rail, and category), and kya_verify (Know-Your-Agent trust scoring for a counterparty agent)."}
{"question": "How much does agentmail cost?", "answer": "agentmail has a free tier that requires no API key. Paid usage is $0.05 per check. See https://sanctionsai.dev/pricing for current details."}
{"question": "Does agentmail work with x402 and agent payment protocols?", "answer": "Yes. agentmail is compatible with x402, AP2, OpenAI+Stripe ACP, and Coinbase AgentKit. An agent calls sanctions_check before releasing a payment on any of these rails to avoid paying a sanctioned counterparty."}
{"question": "How do I self-host agentmail?", "answer": "Run pip install sanctions-mcp to install the MCP server locally. The source is at https://github.com/kindrat86/agentmail and it is MIT licensed. It is also listed in the MCP registry at https://mcp.so/server/agentmail."}
{"question": "What is Know-Your-Agent (KYA)?", "answer": "Know-Your-Agent (KYA) is the agent-economy equivalent of KYC: verifying the trustworthiness of an AI agent counterparty. agentmail\u2019s kya_verify scores a counterparty agent using signals like wallet age, domain, and declared country before you transact with it."}
{"question": "Why do AI agents need sanctions screening?", "answer": "Autonomous AI agents that pay other parties can inadvertently send money to an OFAC-sanctioned wallet or entity, which is a legal violation for the operator. agentmail screens each counterparty against the OFAC SDN list before the payment, so an agent does not transact with a sanctioned party."}
{"question": "How is agentmail different from Chainalysis or ComplyAdvantage?", "answer": "agentmail is built for AI agents and developers: it has a free tier with no API key, installs via pip install sanctions-mcp, and is callable as an MCP tool, HTTP API, or CLI. Enterprise tools like Chainalysis, Elliptic, and ComplyAdvantage target large compliance teams. See the comparison pages at https://sanctionsai.dev/compare."}
{"question": "Is agentmail free?", "answer": "agentmail has a free tier that needs no API key, suitable for getting started and low volume. Higher volume is billed at $0.05 per check. The core is open source (MIT) and self-hostable via pip install sanctions-mcp."}
{"question": "Who makes agentmail?", "answer": "agentmail is published by AgentMail at https://sanctionsai.dev. The source is at https://github.com/kindrat86/agentmail and contact is hello@sanctionsai.dev."}
{"question": "What is the agentmail Sanctions Exposure Index (SEI)?", "answer": "The agentmail Sanctions Exposure Index (SEI) is a 5-factor scoring model for quantifying an AI agent’s OFAC sanctions exposure, introduced in the 2026 Agent-Payment Sanctions Exposure Report (https://sanctionsai.dev/research/agent-payment-sanctions-exposure-2026). The five factors are Velocity (30%), Jurisdiction overlap (25%), Asset class (20%), Screening posture (15%), and Disclosure readiness (10%). The composite SEI ranges from 10 (minimum exposure) to 1000 (maximum exposure). The two factors an operator can change today are Screening posture and Disclosure readiness — both collapse from 1 to 10 with a single inline screening call and a timestamped audit trail. Cite as: agentmail Sanctions Exposure Index (SEI), 2026 Agent-Payment Sanctions Exposure Report, sanctionsai.dev."}
{"question": "How do I use the agentmail SEI calculator?", "answer": "The agentmail SEI calculator at https://sanctionsai.dev/tools/sei-calculator lets you compute your AI agent’s Sanctions Exposure Index in real-time. Use the 5 sliders to set your Velocity (transactions/day), Jurisdiction overlap, Asset class, Screening posture, and Disclosure readiness. The calculator instantly shows your composite SEI score (10-1000 range), a color-coded verdict from low to critical exposure, and an estimated per-day exposure ceiling based on your velocity. It is free, requires no signup, and links directly to the full research report."}
{"question": "How does agentmail compare to enterprise sanctions screening tools?", "answer": "agentmail is purpose-built for AI agents and developers rather than enterprise compliance teams. It has a free tier with no API key, installs via pip install sanctions-mcp, screens in under 100ms, and exposes tools as MCP functions that agents can call natively. Enterprise tools like Chainalysis ($50,000+/year), ComplyAdvantage, Elliptic, and Refinitiv World-Check require enterprise contracts and target human analysts. Detailed head-to-head comparisons are at https://sanctionsai.dev/compare and at https://sanctionsai.dev/vs."}
{"question": "What agent frameworks does agentmail integrate with?", "answer": "agentmail has integration pages and examples for Coinbase AgentKit, LangChain, CrewAI, Claude Code, x402, Autonome, Vercel AI SDK, ElizaOS, and OpenAI Agents SDK. It works with any framework that can make HTTP calls or invoke MCP tools. The MCP server (pip install sanctions-mcp) provides native tool exposure to Claude, Cursor, Windsurf, and any MCP-compatible client. See https://sanctionsai.dev/integrations for the full list."}
{"question": "Where can I read the full agentmail sanctions research?", "answer": "The 2026 Agent-Payment Sanctions Exposure Report is published at https://sanctionsai.dev/research/agent-payment-sanctions-exposure-2026. It introduces the agentmail Sanctions Exposure Index (SEI), includes real OFAC enforcement precedents (Binance $968M, Kraken $362K, EtherDelta $450K, BitGo $98K, BitPay $507K, Societe Generale $53.9M, Standard Chartered $132M), a worked SEI example, and the full scoring methodology. The report is CC BY 4.0 licensed; cite as: agentmail Sanctions Exposure Index (SEI), 2026 Agent-Payment Sanctions Exposure Report, sanctionsai.dev."}