Diliguard MCP Integration | Connect AI Agents

AI Agent Infrastructure // Model Context Protocol

Equip Claude Desktop, Cursor, and AI agents with live corporate intelligence.

Run autonomous KYB registry queries, multi-jurisdiction UBO unwraps, AML sanctions sweeps, and crypto forensics directly from natural language prompts using the official @sluxia/diliguard-mcp package.

Get Your API Key

Explore Agent Skills

How Diliguard MCP works.

The Model Context Protocol (MCP) is an open standard that allows AI models to call external tools and databases directly. Instead of writing custom API integration scripts, you connect the official Diliguard MCP server to your local workspace.

Once connected, your AI assistant can execute complex, multi-source compliance investigations from simple prompts such as:

“Run a complete corporate KYB check on Airbus SE in France, identify all beneficial owners with more than 10 percent equity, and verify whether any directors appear on international sanctions watchlists.”

Local Machine Execution

The MCP server runs locally on your machine via npx. There are no intermediate proxy servers between your AI client and the Diliguard backend.

Secure Environment Variables

Your DILIGUARD_API_KEY is stored securely in your local environment file and passed directly to api.diliguard.com over TLS encryption.

Dynamic Schema Discovery

Your AI model reads tool signatures, parameter schemas, and return structures automatically upon connection.

Available AI tools in @sluxia/diliguard-mcp.

When the Diliguard MCP server is active, your AI assistant gains access to 9 specialized intelligence tools:

Tool Identifier	Capabilities & Data Sources Queried
`run_corporate_kyc`	Queries LEI registry, beneficial ownership across 16 jurisdictions, OFAC/UN sanctions, adverse media, offshore leaks, domain DNS, and VAT status.
`run_person_trace`	Screens individual targets across PEP databases, Interpol notices, court judgments, adverse media, leaked credentials, and email OSINT.
`run_crypto_investigation`	Analyzes cryptocurrency wallet balances, transaction timing, illicit mixer exposure, and cluster counterparty risks across 10+ blockchains.
`run_vendor_risk`	Concurrently screens supplier lists for registry active status, regulatory fines, beneficial ownership changes, and litigation records.
`run_ultimate_fraud`	Executes an end-to-end composite fraud investigation combining corporate entity checks, executive traces, domain footprints, and bank routing validation.
`run_aml_pep_screening`	Performs institutional AML screening across 140+ international sanctions lists, PEP tiers, and regulatory enforcement databases.
`run_registry_check`	Verifies official government registration records, registered office addresses, active filing status, and offshore tax haven links.
`run_ip_intelligence`	Analyzes IP addresses for proxy services, VPNs, Tor exit nodes, hosting data centers, and geographic anomalies.
`run_transaction_scoring`	Evaluates real-time payment transactions by synthesizing counterparty IP intelligence, email reputation, dark web breaches, and IBAN validation.

Setup instructions for Claude Desktop.

To connect Claude Desktop to the Diliguard MCP server, add the configuration block to your local Claude configuration file.

Configuration File Locations

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Configuration JSON Block

{
  "mcpServers": {
    "diliguard": {
      "command": "npx",
      "args": ["-y", "@sluxia/diliguard-mcp"],
      "env": {
        "DILIGUARD_API_URL": "https://api.diliguard.com",
        "DILIGUARD_API_KEY": "YOUR_PERSONAL_API_KEY"
      }
    }
  }
}

After saving the configuration file, restart Claude Desktop. The Diliguard intelligence tools will appear in Claude’s active tool roster.

Testing with the official MCP Inspector.

You can debug your MCP connection, inspect input schemas, and execute test payloads in a local browser UI using the official Model Context Protocol Inspector tool:

DILIGUARD_API_URL="https://api.diliguard.com" DILIGUARD_API_KEY="YOUR_PERSONAL_API_KEY" npx @modelcontextprotocol/inspector npx -y @sluxia/diliguard-mcp

The inspector provides an interactive web interface at http://localhost:5173 where you can trigger tool calls and inspect raw JSON output directly.