Adversarial Corporate Opacity MCP Server

Anti-concealment beneficial ownership detection for AI agents via the Model Context Protocol. This MCP server orchestrates 15 Apify actors across 6 international company registries and 4 sanctions watchlists to deliver corporate ownership graph traversal, cross-lingual sanctions screening with transliteration matching, DBSCAN shell company address clustering, Weisfeiler-Leman infrastructure correlation, loopy belief propagation ownership inference, and weighted opacity scoring with Enhanced Due Diligence (EDD) reporting.

What data can you access?

Data Point	Source	Coverage
Global corporate registries	OpenCorporates	140+ jurisdictions
UK company filings and officers	UK Companies House	All UK companies
Canadian federal corporations	Canada Corp Search	Federal registry
Australian business numbers	Australia ABN Lookup	All ABNs
NZ company registrations	NZ Companies Office	All NZ companies
Legal entity identifiers	GLEIF LEI	Global LEI database
US Treasury sanctions	OFAC Sanctions Search	SDN list
Global sanctions and PEPs	OpenSanctions Search	100+ programs
International wanted persons	Interpol Red Notices	Global notices
US federal wanted list	FBI Most Wanted	FBI database
Domain registration data	WHOIS Lookup	All TLDs
DNS records and configuration	DNS Record Lookup	Any domain
IP geolocation and ASN	IP Geolocation	Global coverage
Certificate transparency logs	crt.sh Search	All certificates
Geographic coordinates	Nominatim Geocoder	Global coverage

MCP Tools

Tool	Price	Description
unfold_ownership_graph	$0.04	BFS traversal of corporate ownership graph across 6 registries with jurisdictional opacity scoring (same 0.1, different 0.3, secrecy 0.5), nominee detection, and formation agent pattern identification.
screen_with_transliteration	$0.04	5-stage cross-lingual name matching against OFAC, OpenSanctions, Interpol, and FBI: Unicode NFKD normalization, Double Metaphone, Caverphone, Jaro-Winkler distance, and token-set ratio with phonetic bonus.
cluster_shell_addresses	$0.04	DBSCAN spatial clustering of registered addresses to detect shell company farms. Identifies locations where many entities share the same or nearby addresses with shell probability scores.
correlate_infrastructure	$0.04	Detect hidden entity relationships via shared digital infrastructure using Weisfeiler-Leman graph kernel similarity. Shared domains, IPs, and TLS certificates reveal common control.
infer_beneficial_owner	$0.04	Infer beneficial owners using loopy belief propagation on a factor graph combining ownership, directorship, address co-location, infrastructure sharing, and sanctions evidence.
compute_entity_opacity_score	$0.04	Weighted opacity scoring from 5 concealment factors: ownership depth, jurisdiction risk, address sharing, sanctions proximity, and infrastructure concealment. Returns LOW/MEDIUM/HIGH/EXTREME.
generate_edd_report	$0.04	Enhanced Due Diligence report aggregating all findings with severity-based risk scoring across ownership structure, sanctions, registration patterns, digital infrastructure, and jurisdictional complexity.

Data Sources

• OpenCorporates -- Global corporate registry covering 140+ jurisdictions with company names, officers, and filing status
• UK Companies House -- UK company registrations, officers, persons with significant control, and filing history
• Canada Corporation Search -- Canadian federal corporate registry with directors and status
• Australia ABN Lookup -- Australian Business Number registry with entity types and GST status
• NZ Companies Office -- New Zealand company registrations, NZBN numbers, and director details
• GLEIF LEI -- Legal Entity Identifier database with parent/child corporate relationships
• OFAC Sanctions Search -- US Treasury SDN list including entity names, aliases, and identification numbers
• OpenSanctions Search -- Consolidated global sanctions, PEPs, and watchlists from 100+ programs
• Interpol Red Notices -- International wanted persons notices
• FBI Most Wanted -- US federal wanted persons database
• WHOIS Lookup -- Domain registration details for digital infrastructure correlation
• DNS Record Lookup -- DNS configuration revealing shared infrastructure and hosting patterns
• IP Geolocation -- IP address mapping for geographic infrastructure analysis
• crt.sh Search -- Certificate transparency logs for TLS certificate relationship discovery
• Nominatim Geocoder -- Address geocoding for spatial clustering analysis

How the scoring works

The server implements five analytical frameworks targeting different concealment techniques.

Ownership Graph Traversal uses BFS across 6 registries with per-hop opacity penalties: 0.1 for same-jurisdiction hops, 0.3 for cross-jurisdiction hops, and 0.5 for secrecy jurisdiction hops. Circular ownership structures are detected and flagged. Nominee directors and formation agent patterns are identified from officer/director name analysis.

Cross-Lingual Transliteration Screening applies a 5-stage pipeline to defeat name evasion: (1) Unicode NFKD normalization with diacritic stripping, (2) Double Metaphone phonetic encoding, (3) Caverphone encoding, (4) Jaro-Winkler distance with prefix bonus, (5) token-set ratio with phonetic bonus when metaphone matches. This catches Cyrillic lookalikes, diacritic evasion, and name reordering.

DBSCAN Address Clustering groups registered addresses spatially to detect shell company farms -- locations where many apparently unrelated entities share the same or nearby addresses. Shell probability scores are computed per cluster.

Weisfeiler-Leman Infrastructure Correlation builds a graph of shared digital assets (domains, IPs, TLS certificates) and computes graph kernel similarity to identify entities under common control despite separate corporate identities.

Loopy Belief Propagation combines all evidence types on a factor graph to infer beneficial ownership probabilities, even when ownership chains are deliberately obscured.

Opacity Level	Score Range	Meaning
LOW	0-1.0	Transparent structure, standard due diligence
MEDIUM	1.0-2.0	Some complexity, enhanced documentation recommended
HIGH	2.0-3.0	Significant concealment indicators, EDD required
EXTREME	3.0+	Multiple concealment techniques detected, escalate

How to connect this MCP server

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "adversarial-corporate-opacity": {
      "url": "https://adversarial-corporate-opacity-mcp.apify.actor/mcp"
    }
  }
}

Programmatic (HTTP)

curl -X POST https://adversarial-corporate-opacity-mcp.apify.actor/mcp \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"unfold_ownership_graph","arguments":{"entity_name":"Acme Holdings Ltd","jurisdiction":"KY"}},"id":1}'

This MCP server also works with Cursor, Windsurf, Cline, and any other MCP-compatible client.

Use cases for corporate opacity intelligence

AML/KYC Beneficial Ownership Verification

Trace multi-layered corporate structures across jurisdictions with unfold_ownership_graph. Detect nominee directors, formation agent patterns, and circular ownership structures that indicate deliberate concealment.

Sanctions Evasion Detection

Screen entities with screen_with_transliteration to catch adversarial name variations that evade standard exact-match screening. The 5-stage pipeline detects Cyrillic lookalikes, diacritic evasion, and name reordering.

Shell Company Farm Detection

Use cluster_shell_addresses to identify registered agent addresses hosting suspiciously many entities. DBSCAN clustering reveals physical address farms used to create the appearance of separate businesses.

Hidden Relationship Discovery

Apply correlate_infrastructure to detect entities sharing digital infrastructure (domains, IPs, TLS certificates) despite having no visible corporate connection. Shared infrastructure often indicates common beneficial ownership.

Enhanced Due Diligence Reporting

Generate comprehensive EDD reports with generate_edd_report for compliance files. Reports aggregate ownership structure, sanctions screening, registration patterns, digital infrastructure, and jurisdictional analysis with severity-based scoring.

Investigative Journalism and Research

Trace offshore corporate structures and identify hidden beneficial owners using the combination of ownership traversal, infrastructure correlation, and belief propagation inference.

How much does it cost?

This MCP server uses pay-per-event pricing at $0.04 per tool call with no subscription fees.

The Apify Free plan includes $5 of monthly platform credits -- enough for 125 tool calls at no cost.

Example costs:

• Ownership graph traversal for a Cayman entity: $0.04
• Cross-lingual sanctions screening for 10 names: $0.04
• Complete 7-tool EDD investigation: $0.28

How it works

1. Your AI agent calls a tool via MCP (e.g., unfold_ownership_graph for an entity)
2. The server dispatches parallel queries to relevant registries and watchlists (3-15 actors depending on tool)
3. Corporate records, sanctions matches, and infrastructure data are collected
4. Mathematical frameworks (BFS traversal, DBSCAN, Weisfeiler-Leman, belief propagation) process the data
5. A structured JSON response is returned with opacity scores, graph structures, and findings

Jurisdiction-specific actors are selected based on the entity's registration. Response time is 30-120 seconds.

FAQ

Q: How many jurisdictions are covered? A: Direct registry access for UK, Australia, Canada, and New Zealand. OpenCorporates extends coverage to 140+ jurisdictions. GLEIF LEI provides global entity identification.

Q: Can this detect all beneficial owners? A: The system infers likely beneficial owners from available public data. Ownership structures using trusts, bearer shares, or jurisdictions with no public registry may not be fully penetrable. The opacity score quantifies how transparent the structure is.

Q: How does transliteration screening differ from standard sanctions matching? A: Standard screening uses exact or basic fuzzy matching. The 5-stage transliteration pipeline catches intentional evasion techniques: diacritic variations, character substitutions, phonetic equivalents, and name reordering.

Q: Is it legal to use this data? A: All data sources are publicly available government registries and open databases. See Apify's guide on web scraping legality.

Q: Does this replace compliance software? A: It provides investigative intelligence to supplement compliance workflows. Regulatory compliance decisions should involve qualified compliance professionals.

Q: Can I run this on a batch of entities? A: Each tool call handles one entity or name list. For batch processing, use the Apify API to run multiple tool calls in parallel.

Related MCP servers

MCP Server	Focus
counterparty-due-diligence-mcp	Broad KYB screening with Counterparty Risk Score
commonwealth-corporate-registry-mcp	Multi-jurisdiction registry search and cross-reference
corporate-political-exposure-mcp	Political exposure and influence mapping

Integrations

This MCP server runs on the Apify platform and supports:

• Scheduling -- Set up recurring ownership and sanctions monitoring via Apify Schedules
• Webhooks -- Trigger alerts when opacity scores change or new sanctions matches appear
• API access -- Call tools directly via the Apify API for compliance system integration
• Dataset export -- Export EDD reports as JSON for audit trail documentation