Run sanctions and PEP screening against the OpenSanctions dataset using the /match API (or self-hosted yente) and tune scoring thresholds to control false positives
domain: opensanctions.org · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Choose between the hosted OpenSanctions API (with a data license or pay-as-you-go plan) and self-hosting the open-source yente API against the same dataset
Submit a structured entity (name, birth date, nationality, entity type) to the /match endpoint rather than a bare text search for better precision
Review the returned scored candidates and their associated risk topics (sanctioned, sanction-linked, PEP, criminal-interest) rather than treating any match as binary
Configure the scoring/matching algorithm parameters to tune sensitivity for your risk appetite, balancing missed hits against false-positive review volume
Re-screen existing customers periodically against updated data, since sanctions and PEP lists change continuously and a point-in-time screen goes stale
Known gotchas
OpenSanctions is free only for non-commercial use; commercial screening requires a data license or paid API subscription even when self-hosting yente against the same data
Name-only matching against a large PEP/sanctions dataset produces significant false positives for common names — always incorporate additional identifiers (DOB, country) to narrow candidates before a human reviews them
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?