D&B Risk Analytics MCP Server

D&B Risk Analytics is part of Dun & Bradstreet's D&B.AI suite of MCP (Model Context Protocol) servers. It exposes the D&B Commercial Graph, anchored by the D-U-N-S Number, so that AI agents can perform verified business identity lookups, evaluate counterparty and supplier exposure, and automate risk and compliance decision documentation. The integration was announced as part of a collaboration with Anthropic to bring KYC/KYB and third-party risk workflows directly inside Claude.

The server provides structured, agent-ready access to D&B's data on more than 600 million public and private businesses across 200+ countries, including firmographics, beneficial ownership, corporate hierarchies, credit and risk scores, and supplier risk signals. Rather than scraping or stitching together raw API responses, the MCP layer delivers verified context at inference time so that downstream agents can reason about ownership, control, and exposure with reduced hallucination risk.

D&B.AI MCP servers are distributed through Databricks Marketplace ("D&B.AI MCP Agent-ready Data") and as remote MCP connectors usable by Claude and other MCP-compatible clients. Access is gated by a D&B commercial entitlement, so the server is targeted primarily at banks, fintechs, insurers, and other regulated enterprises rather than individual developers. Public technical documentation is currently limited, with specific tool names, endpoint URLs, and config schemas typically provided to entitled customers through D&B's developer portal or the Databricks Marketplace listing.

Tools

D&B Risk Analytics MCP is a commercial, provider-hosted server. There is no public GitHub repository or npm package. Access requires a Dun & Bradstreet entitlement plus either a Databricks workspace or an MCP client that supports remote connectors (such as Claude Desktop on a paid plan).

Prerequisites

• An active D&B Risk Analytics or D&B Direct+ subscription with API credentials
• A Databricks workspace (for the Marketplace install path) or a Claude Team/Enterprise plan (for the remote connector path)

Option A: Install via Databricks Marketplace

• In your Databricks workspace, open Marketplace and click "View MCP listings"
• Search for "D&B.AI MCP Agent-ready Data" and click Install
• In the installation dialog, name the Unity Catalog connection (e.g. dnb_risk_analytics)
• The host domain and base path are pre-populated for curated D&B servers
• Provide your D&B API credentials when prompted
• Access is then governed through Unity Catalog permissions and Unity AI Gateway

Option B: Connect as a remote MCP connector in Claude

• In Claude Desktop or Claude.ai (Team/Enterprise), go to Settings, Connectors, Manage Connectors
• Add the D&B remote MCP server URL provided by your D&B account team
• Authenticate with the API key or OAuth credentials issued by D&B

{
  "mcpServers": {
    "dnb-risk-analytics": {
      "url": "https://<your-dnb-mcp-endpoint>",
      "headers": {
        "Authorization": "Bearer ${DNB_API_KEY}"
      }
    }
  }
}

Replace the URL and credentials with the values issued by Dun & Bradstreet. Exact endpoint URLs and tool schemas are distributed to entitled customers and are not published publicly at this time.

• Resolve a counterparty to a verified D-U-N-S Number and pull a full firmographic and risk profile before onboarding
• Map a vendor's corporate family tree and beneficial owners to evaluate concentration and ownership risk across a supplier base
• Monitor third-party suppliers for changes in financial stress, compliance, or operational risk signals and surface alerts to an agent
• Automate KYC/KYB document drafting by combining verified D&B entity data with internal policy rules inside a Claude workflow
• Score a list of new business leads or prospects against D&B credit and viability indicators before sending to sales or underwriting

• "Look up Acme Corp in the D&B Commercial Graph and return the D-U-N-S Number, headquarters address, and ultimate parent."
• "For D-U-N-S 123456789, pull the latest risk and viability scores and summarize key drivers."
• "Build the corporate family tree for this supplier and flag any subsidiaries in sanctioned jurisdictions."
• "Generate a KYB onboarding memo for this prospective customer using verified D&B data and our internal risk policy."
• "Monitor my supplier list and alert me to any companies whose financial stress score has worsened in the last 30 days."

• Backed by D&B's authoritative Commercial Graph covering 600M+ businesses, with D-U-N-S Number as a stable global identifier
• Official, provider-hosted server maintained by Dun & Bradstreet rather than a community fork
• Distributed through Databricks Marketplace with Unity Catalog and AI Gateway governance for enterprise control
• Designed for regulated KYC/KYB and third-party risk workflows, including audit-ready documentation generation

• Requires a paid D&B subscription and entitlement; not usable for individual developers or evaluation without a commercial agreement
• Public technical documentation is sparse: specific tool names, endpoint URLs, and schemas are not published openly and must be obtained from D&B
• No open-source repository, so behavior and tool surface cannot be inspected or self-hosted

• Moody's MCP server on Databricks Marketplace for credit and entity risk data
• S&P Global MCP server for financial and counterparty intelligence
• D&B Direct+ REST API used directly (without MCP) via a custom MCP wrapper for read-only firmographic and risk queries