Sanctions Evasion Network MCP Server

Sanctions evasion detection that goes beyond name matching — this MCP server traces ownership chains across 8 corporate and sanctions databases, maps director networks for nominee patterns, and produces an Evasion Probability Score (0–100) for any entity you submit. Built for compliance teams, due diligence analysts, and trade finance professionals who need structural insight, not just list lookups.

Connect it to Claude Desktop, Cursor, Windsurf, or any MCP-compatible client and screen counterparties, trace beneficial owners, and assess trade route risk directly from your AI workflow. Seven specialist tools run up to 8 data actors in parallel, so a full deep screen across OFAC, OpenSanctions, Interpol, OpenCorporates, UK Companies House, GLEIF LEI, Canada Corporations, and Australia ABN completes in a single tool call.

What data can you access?

Why use Sanctions Evasion Network MCP Server?

Standard sanctions screening fails at one specific and well-documented problem: sanctioned parties rarely appear on lists under their own name. They use layered corporate structures — shell companies in the British Virgin Islands owning holding companies in Luxembourg owning operating companies in transparent jurisdictions — to distance themselves from SDN entries. A name-match tool gives you a clean result, and you proceed. That is the gap this MCP closes.

This server queries 8 live databases in parallel, then runs 4 scoring models — ownership chain depth, director network analysis, jurisdictional risk cascade, and multi-database hit correlation — before combining them into a single Evasion Probability Score. The result tells you not just whether the name matched, but whether the corporate structure looks designed to hide who is really in control.

• Scheduling — run daily counterparty monitoring passes on client portfolios to catch new sanctions designations
• API access — trigger entity screens from Python, JavaScript, or any HTTP client within your existing compliance pipeline
• Pay-per-event pricing — pay $0.045 per tool call, with no monthly subscription and a configurable spending cap per session
• Monitoring — get Slack or email alerts when Evasion Probability Scores exceed defined thresholds via Apify webhooks
• Integrations — connect to Zapier, Make, or your case management system to route flagged entities into a review workflow automatically

Features

• 7 specialist MCP tools covering the full sanctions evasion detection workflow: from quick sanctions list matching to full composite structural screening
• 8 parallel data actors called simultaneously per deep_entity_screening query — OFAC, OpenSanctions, Interpol, OpenCorporates, UK Companies House, GLEIF LEI, Canada Corporations, Australia ABN
• Evasion Probability Score (0–100) computed as a weighted composite: sanctions hits 35%, ownership chain 25%, jurisdictional risk 25%, director network 15%
• FATF blacklist cascade — 5 blacklisted jurisdictions (DPRK, Iran, Myanmar, and others) trigger a 15-point penalty per hit, with an additional amplifier when combined with secrecy jurisdictions
• FATF greylist coverage — 26 greylisted jurisdictions scored at 8 points each, flagging enhanced due diligence requirements
• 18 secrecy jurisdictions indexed — including BVI, Cayman Islands, Panama, Seychelles, Marshall Islands, Liechtenstein, Luxembourg, and Mauritius
• Shell company indicator engine — scores entities on officer count, filing history, company age, registered agent address patterns, and entity names containing "Holdings", "International", "Ventures"
• Director network cross-directorship detection — counts individuals appearing across multiple entities; individuals on 5+ entities receive a 3x amplifier
• Address clustering analysis — detects shared registered agent addresses across entity networks, a strong indicator of nominee-serviced shell structures
• Rapid incorporation pattern detection — flags clusters of entities incorporated within 30-day windows as a structural evasion indicator
• GLEIF LEI chain traversal — checks for lapsed or retired LEI registrations and missing LEI assignments; absence of LEI in a corporate chain adds 15 points to ownership chain score
• Multi-database hit correlation — bonus scoring when the same entity appears across OFAC, OpenSanctions, and Interpol simultaneously
• Alias and transliteration support — fuzzy_sanctions_match accepts known aliases to broaden query coverage across Cyrillic, Arabic, and Chinese romanization variants
• Verdict thresholds with override rules — CONFIRMED_HIT or any Interpol red notice forces BLOCK_TRANSACTION regardless of composite score
• Structured recommendation output — each report includes actionable recommendations: SAR filing guidance, beneficial ownership disclosure requirements, OFAC license flags
• Spending limit enforcement — all 7 tools check charge limits before executing, preventing overrun in automated pipelines

Use cases for sanctions evasion detection

Trade finance counterparty screening

Banks and commodity traders must screen counterparties before approving letters of credit or trade finance instruments. A counterparty may have a clean SDN name-match result but be beneficially owned by a sanctioned Russian oligarch through a BVI holding structure. The trade_route_risk_assessment tool runs sanctions screening alongside ownership chain and jurisdictional risk analysis in a single call, returning a verdict and structured recommendations the compliance officer can attach to the transaction file.

Corporate KYB for financial institutions

Correspondent banks and payment processors conducting Know Your Business verification need to trace beneficial ownership to the 25% threshold required by FinCEN. The beneficial_owner_identification tool searches OpenCorporates, UK Companies House, and GLEIF simultaneously to surface the ownership chain, then screens discovered individuals against OFAC and OpenSanctions in the same call.

Enhanced due diligence for private equity and M&A

Deal teams screening acquisition targets or fund investors face complex offshore structures. A target headquartered in the UK may have ultimate ownership in jurisdictions that trigger FATF concerns. The ownership_chain_trace tool maps the ownership structure across 5 corporate registries, scoring corporate nesting depth, secrecy jurisdiction count, and shell indicators, then returns a TRANSPARENT / MINOR_COMPLEXITY / OPAQUE / HIGHLY_LAYERED / EVASION_PATTERN classification.

Compliance workflow automation for AI agents

Compliance automation teams building AI-powered case management can embed this MCP server into Claude or GPT-based agents. The agent calls deep_entity_screening when processing new counterparty onboarding requests, receives structured JSON with score, verdict, signals, and recommendations, and routes cases above a configurable threshold to human review — no manual screening queue required.

Sanctions evasion investigation and forensics

Financial intelligence units and investigative journalists tracing oligarch asset structures can use director_network_map to uncover nominee director networks — clusters of individuals serving as directors across dozens of shell companies registered at the same Cayman Islands registered agent address. Address clustering scores and cross-directorship counts surface these patterns quantitatively.

Export control pre-screening

Export compliance officers verifying end-user statements before shipping controlled technology can use jurisdictional_risk_score to flag entities with connections to FATF blacklisted or greylisted jurisdictions before a full export licence determination, reducing the manual research burden in high-volume shipment pipelines.

How to screen entities for sanctions evasion

1. Connect the MCP server to your AI client — add the server URL https://sanctions-evasion-network-mcp.apify.actor/mcp to your Claude Desktop, Cursor, or Windsurf MCP configuration with your Apify API token as the Bearer credential.
2. Ask your AI to screen an entity — type a natural language instruction such as "Screen Pinnacle Global Resources Ltd for sanctions exposure and ownership red flags." The AI calls deep_entity_screening and receives the full structured report.
3. Review the Evasion Probability Score and signals — the response includes a 0–100 composite score, a five-tier verdict (CLEARED / LOW_RISK / ENHANCED_DUE_DILIGENCE / HIGH_RISK / BLOCK_TRANSACTION), per-dimension scores, specific signal strings explaining each flag, and action recommendations.
4. Act on the recommendations — the report includes specific next steps: request full beneficial ownership disclosure, engage your compliance team, file a SAR, or obtain an OFAC licence. For CLEARED verdicts, retain the JSON response as your screening documentation.

MCP tools

Tool input parameters

Output example

A deep_entity_screening call for a high-risk entity returns:

{
  "entity": "Meridian Commodities Group Ltd",
  "compositeScore": 74,
  "verdict": "HIGH_RISK",
  "ownershipChain": {
    "score": 68,
    "nestingDepth": 4,
    "secrecyJurisdictions": 3,
    "shellIndicators": 6,
    "chainLevel": "HIGHLY_LAYERED",
    "signals": [
      "3 secrecy jurisdictions — layered offshore structure",
      "6 shell indicators — nominee/dormant company patterns",
      "No LEI registration — entity lacks transparent identifier",
      "Entity spans 5 jurisdictions — complex cross-border structure"
    ]
  },
  "directorNetwork": {
    "score": 52,
    "nomineePatterns": 3,
    "crossDirectorships": 4,
    "networkLevel": "SUSPICIOUS",
    "signals": [
      "4 individuals serve on multiple entities — network pattern",
      "3 nominee/agent roles detected — potential front persons",
      "2 shared registered addresses — shell company cluster"
    ]
  },
  "jurisdictionalRisk": {
    "score": 41,
    "fatfBlacklist": 0,
    "fatfGreylist": 3,
    "secrecyJurisdictions": 3,
    "riskLevel": "ELEVATED",
    "signals": [
      "3 FATF greylist jurisdictions — enhanced due diligence required",
      "3 secrecy jurisdictions — financial opacity risk"
    ]
  },
  "sanctionsHits": {
    "score": 28,
    "ofacHits": 1,
    "openSanctionsHits": 2,
    "interpolHits": 0,
    "hitLevel": "PARTIAL_MATCH",
    "signals": [
      "Appears on 2 sanctions lists — multi-jurisdictional exposure"
    ]
  },
  "allSignals": [
    "Appears on 2 sanctions lists — multi-jurisdictional exposure",
    "3 secrecy jurisdictions — layered offshore structure",
    "6 shell indicators — nominee/dormant company patterns",
    "No LEI registration — entity lacks transparent identifier",
    "Entity spans 5 jurisdictions — complex cross-border structure",
    "4 individuals serve on multiple entities — network pattern",
    "3 nominee/agent roles detected — potential front persons",
    "3 FATF greylist jurisdictions — enhanced due diligence required"
  ],
  "recommendations": [
    "Sanctions exposure detected — engage compliance team before proceeding",
    "File Suspicious Activity Report (SAR) — multiple risk indicators triggered"
  ]
}

Output fields

Verdict reference

Override rules: A CONFIRMED_HIT sanctions level or any Interpol Red Notice result forces BLOCK_TRANSACTION regardless of the composite score.

How much does it cost to screen entities for sanctions evasion?

This MCP uses pay-per-event pricing — you pay $0.045 per tool call. Platform compute costs are included. There is no monthly subscription fee.

You can set a maximum spending limit per session to control costs. Each tool checks the charge limit before executing and returns a clear error if the limit is reached.

Compare this to enterprise KYC/sanctions platforms that charge $2,000–15,000 per month for comparable structural evasion analysis — with this MCP, most compliance teams running 50–200 screens per month spend under $10, with no subscription commitment and no minimum contract.

The Apify Free plan includes $5 of monthly platform credits — enough for approximately 100 tool calls with no payment required.

How to connect this MCP server

Claude Desktop

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "sanctions-evasion-network": {
      "url": "https://sanctions-evasion-network-mcp.apify.actor/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Python (direct API)

import requests

response = requests.post(
    "https://sanctions-evasion-network-mcp.apify.actor/mcp",
    headers={
        "Content-Type": "application/json",
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
    },
    json={
        "jsonrpc": "2.0",
        "method": "tools/call",
        "params": {
            "name": "deep_entity_screening",
            "arguments": {
                "entity": "Meridian Commodities Group Ltd"
            }
        },
        "id": 1
    }
)

result = response.json()
report = result["result"]["content"][0]["text"]
import json
data = json.loads(report)
print(f"Entity: {data['entity']}")
print(f"Composite score: {data['compositeScore']}/100")
print(f"Verdict: {data['verdict']}")
for signal in data["allSignals"]:
    print(f"  - {signal}")
for rec in data["recommendations"]:
    print(f"  => {rec}")

JavaScript

const response = await fetch("https://sanctions-evasion-network-mcp.apify.actor/mcp", {
    method: "POST",
    headers: {
        "Content-Type": "application/json",
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
    },
    body: JSON.stringify({
        jsonrpc: "2.0",
        method: "tools/call",
        params: {
            name: "fuzzy_sanctions_match",
            arguments: {
                name: "Viktor Semenov",
                aliases: "Victor Semenoff, В. Семенов"
            }
        },
        id: 1
    })
});

const result = await response.json();
const report = JSON.parse(result.result.content[0].text);
console.log(`Sanctions hit level: ${report.sanctionsHits.hitLevel}`);
console.log(`OFAC hits: ${report.sanctionsHits.ofacHits}`);
console.log(`OpenSanctions hits: ${report.sanctionsHits.openSanctionsHits}`);
console.log(`Interpol hits: ${report.sanctionsHits.interpolHits}`);

cURL

# Deep entity screen
curl -X POST "https://sanctions-evasion-network-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "deep_entity_screening",
      "arguments": { "entity": "Meridian Commodities Group Ltd" }
    },
    "id": 1
  }'

# Fuzzy sanctions match with aliases
curl -X POST "https://sanctions-evasion-network-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "fuzzy_sanctions_match",
      "arguments": {
        "name": "Petrocom Trading LLC",
        "aliases": "Petrochem Trading, ООО Петроком"
      }
    },
    "id": 2
  }'

How Sanctions Evasion Network MCP works

Parallel data collection across 8 sources

When a tool call arrives, the server calls between 3 and 8 Apify actors simultaneously using Promise.all. Each actor runs with a 512 MB memory allocation and a 120-second timeout. OFAC Sanctions, OpenSanctions, and Interpol Red Notices are queried for the entity name. OpenCorporates (140+ jurisdictions), UK Companies House, GLEIF LEI, Canada Corporations, and Australia ABN are queried for corporate registry data. If any individual actor fails or times out, it returns an empty array rather than propagating the error, ensuring the other data sources still contribute to the score.

Four-dimensional scoring

The scoreOwnershipChain function aggregates all corporate records from the 5 registries and computes a score across four components: nesting depth (max 30 points), shell company indicators using officer count, filing history, company age, and entity name patterns (max 25 points), GLEIF LEI chain depth and missing-LEI penalty (max 25 points), and multi-jurisdiction complexity with an 8-point amplifier for entities spanning 4+ jurisdictions (max 20 points). Shell indicators include officer count ≤ 1, zero filings after 1 year of existence, and names containing "Holdings", "International", "Group", or "Ventures".

The scoreDirectorNetwork function builds a director frequency map from officer arrays across all corporate records and counts individuals appearing on 2+ entities as cross-directorships. An individual on 5+ entities adds 3 bonus cross-directorship units. Address clustering runs a separate frequency map on registered addresses, scoring shared addresses at 8 points each. Rapid incorporation detection sorts all entity incorporation dates and flags pairs created within 30 days of each other.

The scoreJurisdictionalRisk function extracts jurisdiction strings from every corporate record field (jurisdiction, country, legalAddress.country, registeredOffice), normalises to lowercase, and scores against three hard-coded lists: 5 FATF blacklisted countries at 15 points each (max 35), 26 FATF greylisted countries at 8 points each (max 25), and 18 secrecy jurisdictions at 6 points each (max 25). A cascade amplifier adds up to 15 points when blacklist and secrecy hits co-occur or when 2+ greylist hits are present.

The scoreSanctionsHits function scores OFAC results at 10 points for strong matches (score ≥ 90 or type including "individual" or "entity") plus 3 points per result (max 35). OpenSanctions results are scored by unique sanctions list count at 3 points each plus 4 points per result (max 30). Interpol hits score at 10 points each (max 20). A cross-database correlation bonus of 5 points per database with any hit (max 15) rewards entities appearing across multiple systems.

Composite scoring and verdict assignment

The generateEvasionReport function combines the four sub-scores with weights: sanctions hits × 0.35, ownership chain × 0.25, jurisdictional risk × 0.25, director network × 0.15. Verdict thresholds are: CLEARED below 15, LOW_RISK 15–34, ENHANCED_DUE_DILIGENCE 35–54, HIGH_RISK 55–74, BLOCK_TRANSACTION 75 and above. Two override rules force BLOCK_TRANSACTION regardless of score: a CONFIRMED_HIT sanctions level (score ≥ 80) or any Interpol Red Notice result. Recommendations are generated conditionally: SAR filing is recommended when composite score is 55+; OFAC licence guidance when any FATF blacklisted jurisdiction is found; beneficial ownership disclosure when chainLevel is EVASION_PATTERN.

Tips for best results

1. Use deep_entity_screening for new counterparties and fuzzy_sanctions_match for periodic re-checks. Full screening costs the same as a narrow match but provides the ownership chain and director network signals needed for enhanced due diligence documentation. Re-checks on already-cleared entities can use fuzzy_sanctions_match for a faster, more focused result.

2. Provide aliases in fuzzy_sanctions_match for individuals with Cyrillic, Arabic, or Chinese names. The tool concatenates the alias string with the primary name before querying all three databases. Cover common romanization variants (e.g., Prokhorov / Prokhoroff) and patronymic forms to maximise recall.

3. Set a spending limit when running bulk screening batches. Call the MCP server from an automated script with a cap of $5 or $10 to prevent runaway costs if your counterparty list is larger than expected. The server returns a structured error when the limit is reached so your pipeline can pause gracefully.

4. Use ownership_chain_trace to support beneficial ownership documentation. The tool returns raw corporate records from up to 15 OpenCorporates results and 10 GLEIF records alongside the chain score. Export these as JSON and attach to the KYB case file.

5. Interpret chainLevel: OPAQUE or higher as a trigger for enhanced due diligence, not an automatic block. Complex ownership structures are common in legitimate multinational businesses. The Evasion Probability Score contextualises complexity with sanctions exposure and director risk — use all four dimensions together.

6. Combine with other MCP servers for full financial crime coverage. Pair this server with an AML/CFT screening server for transaction monitoring context, or with an export control screening server for dual-use goods compliance. The structured JSON output integrates directly with any MCP-capable AI agent workflow.

7. Run director_network_map when investigating an individual rather than a company. You can submit a person's name as the entity parameter to surface all corporate entities associated with that director across OpenCorporates, UK Companies House, and Canada Corporations.

Combine with other Apify actors

Limitations

• Name matching quality depends on spelling consistency. Corporate registries use inconsistent name formats. Abbreviated names, transliterated names, and entities registered under trading names different from their official name may not surface in registry searches. Always screen with known aliases.
• Private ownership beyond registered filings is not accessible. Beneficial ownership held through trusts, nominee share structures, or unregistered agreements is invisible to registry-based analysis. The score reflects structural indicators of opacity, not confirmed proof of hidden ownership.
• Data currency varies by registry. Corporate registry filings reflect the last submitted update. UK Companies House data is updated in near-real time; some OpenCorporates jurisdictions have multi-month lag in filing ingestion.
• The composite score is probabilistic, not deterministic. A score of 75+ indicates a high probability that the structure was designed to obscure ownership — it does not confirm a sanctions violation. Human compliance judgement is required before taking consequential action.
• FATF and secrecy jurisdiction lists are hard-coded at build time. If FATF updates its grey or blacklists between builds, the scoring will not reflect the update until the actor is rebuilt. Monitor FATF announcements for list changes.
• The actor does not screen financial transactions. It screens entities and corporate structures. For transaction-level screening (wire transfers, invoice counterparties), the output of this MCP must be embedded into a transaction monitoring workflow.
• Interpol Red Notice search is based on the public Interpol website. Diffuse notice data (restricted notices, domestic warrants) is not available through this source.
• Standby mode requires an active Apify account with a paid or free plan. The server runs persistently in Apify's Standby infrastructure. Cold start after inactivity takes approximately 5–15 seconds for the first request.

Integrations

• Claude Desktop — add the MCP server URL to claude_desktop_config.json and screen entities conversationally
• Cursor / Windsurf / Cline — integrate into your development environment for sanctions screening during code-assisted compliance workflows
• Apify API — call the MCP endpoint programmatically from any language using standard HTTP POST requests
• Webhooks — trigger downstream case management workflows when evasion scores exceed configurable thresholds
• Zapier — route high-risk screening results into Salesforce, HubSpot, Jira, or email alert workflows without code
• Make — build multi-step compliance automation scenarios connecting screening results to case management systems
• LangChain / LlamaIndex — embed this MCP server as a tool in autonomous compliance agent pipelines

Troubleshooting

• Spending limit reached error on the first call — the per-session spending limit may be set too low. Check the limit configuration in your MCP client or Apify account. Each tool call costs $0.045 so a limit of $1 allows approximately 22 calls.

• Score is 0 despite knowing the entity has sanctions exposure — the query name may not match the exact registered name in the source databases. Try fuzzy_sanctions_match with known aliases or transliterated variants. Also check that the entity name is spelled as it appears in OFAC SDN listings.

• ownership_chain_trace returns empty corporate records — the entity may be registered under a different jurisdiction name than expected, or may not have a presence in the 5 registries queried. Try opencorporates-search directly with alternate name spellings. Entities registered only in non-English-alphabet jurisdictions (China, Russia, Gulf states) may not surface in these registries.

• Slow response on the first request — this MCP runs in Apify Standby mode. If the server has been idle, the first request incurs a cold start of 5–15 seconds. Subsequent requests within the same session respond immediately. For latency-sensitive pipelines, send a lightweight health-check call before the production query.

• Director network signals not appearing for a known network — director_network_map relies on officer arrays in corporate registry responses. If the registries searched do not include officer data for the queried entity (e.g., Australian ABN records have limited officer data), cross-directorship detection will undercount. Add the entity's UK or Canadian registration number to the query for richer officer data.

Responsible use

• All data queried by this MCP server is sourced from publicly available government databases, international regulatory bodies, and open corporate registries.
• Screening results are probabilistic indicators, not legal determinations. Do not use Evasion Probability Scores as the sole basis for blocking transactions, denying services, or making public accusations.
• Comply with applicable data protection regulations (GDPR, CCPA) when processing personal data returned in director and individual screening results.
• Do not use screening results to discriminate based on national origin, ethnicity, or religion. FATF jurisdiction risk scoring reflects government-assessed financial crime risk, not individual culpability.
• For guidance on the legality of accessing public registry data, see Apify's guide on web scraping legality.

FAQ

How is Sanctions Evasion Network MCP different from standard sanctions screening tools? Standard tools match names against sanctions lists. This MCP adds four structural analysis layers — ownership chain depth, director network mapping, jurisdictional risk cascade, and multi-database hit correlation — to detect entities that use corporate complexity to distance themselves from SDN designations. It is designed to catch the structures that name matching misses.

How does sanctions evasion detection work when the entity name is clean on the SDN list? The server queries 5 corporate registries to build an ownership chain and director network map. If that chain runs through secrecy jurisdictions, uses nominee directors, contains shell company indicators, or spans multiple FATF-risk jurisdictions, the Evasion Probability Score rises even with zero direct sanctions hits. A score above 55 with a clean SDN result still warrants enhanced due diligence.

Can I screen individuals as well as companies? Yes. All 7 tools accept person names. For individuals, fuzzy_sanctions_match and deep_entity_screening are most effective. director_network_map will surface all companies where the individual appears as a director or officer across OpenCorporates, UK Companies House, and Canada Corporations.

How accurate is the Evasion Probability Score? The score quantifies structural risk indicators derived from publicly available data. It is calibrated to flag evasion-pattern structures with high sensitivity, which means there is a false positive rate for legitimately complex multinational businesses. Treat scores of 55–74 (HIGH_RISK) as triggers for manual review, not automatic rejection. Scores of 75+ should be escalated to your compliance team.

How many sanctions lists does the fuzzy_sanctions_match tool cover? The tool queries OFAC SDN (US Treasury), OpenSanctions (which aggregates EU, UN, UK, OFAC, and 40+ additional national lists), and Interpol Red Notices. An entity appearing on 2+ distinct lists within OpenSanctions receives a corroboration bonus in the scoring model.

Can sanctions evasion detection be automated for ongoing counterparty monitoring? Yes. Use Apify's scheduling feature to run periodic screening batches, or integrate the MCP endpoint into your CRM or counterparty management system via webhooks. When an entity's score crosses a threshold between screening runs, the webhook triggers an alert to your compliance team.

What does BLOCK_TRANSACTION mean in practice? The BLOCK_TRANSACTION verdict is generated when the composite score is 75 or above, when sanctionsHits.hitLevel reaches CONFIRMED_HIT, or when any Interpol Red Notice is returned. It is a recommendation to halt the transaction pending a full compliance investigation — not an automated block. Your compliance system must implement the blocking action.

Is it legal to screen entities using public registry and sanctions data? Yes. All data sources are publicly available government and international databases: OFAC (US Treasury), OpenSanctions (aggregated public lists), Interpol (public Red Notices), OpenCorporates, UK Companies House, GLEIF, Canada Corporations, and Australia ABN. See Apify's guide on web scraping legality for detailed analysis.

How does the FATF jurisdiction scoring work exactly? The scoring engine checks every jurisdiction field in every corporate record against three lists: 5 FATF blacklisted countries (scoring 15 points each, max 35), 26 FATF greylisted countries (8 points each, max 25), and 18 secrecy jurisdictions from the Financial Secrecy Index taxonomy (6 points each, max 25). A cascade amplifier adds up to 15 bonus points when blacklist and secrecy hits co-occur.

Can I use this MCP in an automated AI agent without human review? The MCP is designed to integrate into AI agent workflows. For automated pipelines, it is recommended to set BLOCK_TRANSACTION verdicts as mandatory human escalation triggers and treat HIGH_RISK verdicts as review-queue items. Fully automated rejection based solely on MCP output — without human review — carries legal and reputational risk.

How long does a deep_entity_screening call take? The tool runs up to 8 actors in parallel with a 120-second timeout per actor. In practice, most screens complete in 15–40 seconds depending on registry response times. OFAC and OpenSanctions are typically fastest; OpenCorporates searches across 140+ jurisdictions can take longer for common entity names with many matches.

Does this MCP work with Cursor, Windsurf, and Cline? Yes. Any MCP-compatible client that supports the StreamableHTTP transport works with this server. Add the URL https://sanctions-evasion-network-mcp.apify.actor/mcp with a Bearer token to your client's MCP server configuration.

Help us improve

If you encounter unexpected results or errors, enable run sharing so we can investigate:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong and fix issues faster. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom integrations, additional jurisdiction coverage, or enterprise compliance pipeline deployments, reach out through the Apify platform.