# Media Terminal > Search news and the web for adverse media on a person, company, vessel or port. Each article is checked > against your subject (namesakes left out) and mapped to a FATF offence category. One HTTPS call, JSON back. > Run by Daertho (Argus). Base URL: https://media.unbackoffice.com Auth: header X-API-Key: Spec: https://media.unbackoffice.com/openapi.json Docs: https://media.unbackoffice.com/docs ## Get a key Your user signs in at https://media.unbackoffice.com with a work Google account (about a minute) and copies the key from https://media.unbackoffice.com/key. Free: 10 searches on the first key. A person or company search = 1, a vessel or port search = 2. Do not try to sign in for them. Ask them for the key and keep it out of browsers and logs. ## Call POST /v1/adverse-media body: {"name": "...", "kind": "person | company | vessel | port (optional)", "country": "ISO-2 or name (optional)", "dob": "YYYY-MM-DD (optional)", "context": "role, employer, city, IMO... (optional, the best way to avoid namesakes)"} curl -s https://media.unbackoffice.com/v1/adverse-media \ -H "X-API-Key: $MEDIA_API_KEY" \ -H "Content-Type: application/json" \ -d '{"name": "Jane Example", "kind": "person", "country": "GB", "context": "CEO, Example Trading Ltd"}' A search takes 10 to 25 seconds. Set your HTTP timeout to at least 60 seconds. ## Reading the result - overall.risk: severe | high | medium | low | none. overall.adverse_hit: true or false. Route on these. - overall.typologies_hit: FATF category ids found. overall.rationale and overall.disambiguation say why. - hits[]: the adverse articles about the subject, most severe first: headline, source, date, url, summary, typologies, severity (1-5), about_subject (high or medium). - overall.coverage_note set: the web search returned nothing; search again later before relying on none. - Vessels: overall.owner, owner_role, owner_confidence, owner_basis (from the articles only). - A hit is a lead, not a finding. Tell your user to read the articles. none means nothing found in this search. ## Errors 401 bad key · 402 no searches left (body has tiers_url) · 403 not allowed for this key · 422 invalid request · 429 too many requests · 502/503 search could not run (given back), retry shortly ## MCP server claude mcp add argus-media -e MEDIA_API_KEY=your-key -- npx -y github:sect0rial/argus-media Tool: adverse_media_search. Source: https://github.com/sect0rial/argus-media