io.github.qinisolabs/sanctionwise

sanctionwise

Verified UK sanctions screening for AI agents — official FCDO UK Sanctions List, not guesses.

Verified, trustworthy data tools for AI agents. "Qiniso" means "truth" in Zulu.

Website · npm · MCP Registry

Ask an LLM "is this person on the UK sanctions list?" and it will answer from stale, fuzzy memory. Designations change constantly and are specific — that's data, not something to recall. sanctionwise screens a name against the official FCDO UK Sanctions List and returns possible matches with their real designation details, or a clean "no match".

⚠️ Read this — what this is and is NOT

• Indicative name-match only. A match is a POSSIBLE match that a human must verify against the official entry (names are widely shared and transliterated). It is not confirmation that the party you mean is the sanctioned party.
• A "no match" is NOT a clearance. It covers the UK (FCDO) list only, screens the name string you pass (not DOBs or aliases you didn't supply), and reflects the dataset's report date, not real-time.
• Not advice. Not legal, compliance, KYC/AML or sanctions advice, and not a substitute for a regulated sanctions-screening process. Every response includes this disclaimer.

Install

{ "mcpServers": { "sanctionwise": { "command": "npx", "args": ["-y", "sanctionwise"] } } }

The official UK Sanctions List data is bundled in the package (works offline, out of the box). The tool also auto-refreshes from a weekly-updated GitHub Release on startup — cached locally, no key, no token, no manual step — so it stays current on its own. Every response reports the dataset's report date.

Use it as a library

npm i sanctionwise

import { screenName, getSanctionsEntry } from "sanctionwise";

screenName("Vladimir Putin");
// { matchCount: 3, results: [{ id: "RUS0251", type: "Individual",
//   primaryName: "Vladimir Vladimirovich PUTIN", matchType: "exact",
//   regimes: ["The Russia (Sanctions) (EU Exit) Regulations 2019"], sanctions: "Asset freeze|…" }],
//   disclaimer: "INDICATIVE NAME-MATCH SCREEN ONLY. …" }

screenName("Acme Quilting Supplies Ltd").matchCount;   // 0 — no match (NOT a clearance)
getSanctionsEntry("RUS0251");                          // full official entry for a Unique ID

Matching folds accents and is case/punctuation-insensitive, and screens primary names and aliases.

Tools — 2

Data & auto-refresh

The data is the official FCDO UK Sanctions List (designated persons, entities and ships under the Sanctions and Anti-Money Laundering Act 2018), published under the Open Government Licence v3.0. scripts/build-data.mjs parses the official CSV into one record per target (all names/aliases grouped); the bundled build currently carries 6,200+ designated targets. A weekly GitHub Action (.github/workflows/refresh-data.yml) re-downloads the list and publishes it as a GitHub Release asset; the installed tool fetches that asset on startup (cached, offline-safe), so it self-updates with no npm token and no manual step. The bundled copy is the offline / first-run fallback.

What it is not

• Not a compliance determination. Indicative screening to surface possible matches for human review — nothing more.
• Not multi-list. UK (FCDO) list only — not the EU, US (OFAC), UN consolidated, or any other list.
• Not real-time. Reflects the dataset's report date; check the live list for the current position.
• Not advice, and not a regulated screening service.

Privacy

This tool runs locally on your machine and is built not to collect, store, or transmit your data — no analytics, no telemetry, no account. Screening runs locally; on startup it may download the latest public UK Sanctions List from a GitHub Release, but the names you screen are never transmitted. Full policy: https://qinisolabs.github.io/privacy.html.

License

Apache-2.0 (code). Sanctions data © Crown copyright, FCDO, Open Government Licence v3.0; see NOTICE.