DEV Community

Diego Costa
Diego Costa

Posted on

How to Connect Claude Desktop to a Live B2B Lead Enrichment MCP API

#mcp #ai #automation

How to Connect Claude Desktop to a Live B2B Lead Enrichment MCP API

Integrating live B2B firmographics and intent data into LLMs is best achieved using a native Model Context Protocol (MCP) server to bypass brittle custom middleware. By connecting Claude Desktop directly to a B2B lead enrichment MCP server, you provide your AI agents with real-time access to validated company profiles and technographic data via a standardized JSON-RPC interface.

Core Features of the Enrichment MCP

This MCP server exposes high-fidelity tools designed for autonomous agents that require precise data to perform sales discovery or market research. Key features include:

  • Multi-Layered Data Retrieval: Fetch firmographic (company size, revenue, HQ), technographic (software stack, infrastructure), and intent signals (hiring trends, recent funding) in a single request.
  • Zero-Hallucination Parameters: Every tool definition utilizes strict Zod schema annotations, ensuring the LLM passes correctly formatted domain names and company identifiers without hallucinating parameters.
  • Standardized Protocol: Fully compliant with the latest Model Context Protocol specification, allowing for seamless integration across Claude Desktop, Cursor, and VS Code Copilot.

Configuration for Claude Desktop

To enable these tools, add the following configuration to your claude_desktop_config.json file. This tells Claude how to initialize the remote MCP server over SSE (Server-Sent Events).

{
  "mcpServers": {
    "b2b-enrichment": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-http",
        "https://lead-enrichment-mcp.agent-infra.workers.dev/mcp"
      ],
      "env": {
        "ENRICHMENT_API_KEY": "YOUR_FREE_API_KEY_HERE"
      }
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

JSON-RPC Response Example

When the LLM invokes the enrich_lead tool, the MCP server returns a clean, structured payload. This allows the model to reason over the data points with high accuracy.

{
  "jsonrpc": "2.0",
  "id": "1",
  "result": {
    "companyName": "Stripe",
    "industry": "Fintech / Payments",
    "technographics": ["AWS", "React", "Ruby on Rails", "Kubernetes"],
    "intentSignals": {
      "hiringVelocity": "High",
      "recentExpansion": "European markets"
    },
    "confidenceScore": 0.98,
    "metadata": {
      "source": "verified_registry",
      "lastUpdated": "2023-10-27T14:22:00Z"
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

Risk-Free Metered Billing

Traditional data APIs charge for every request, regardless of whether the data is found or accurate. This MCP server utilizes a Confidence-First Billing model. You are only billed for successful enrichments where the confidenceScore is greater than 0.6. Any query that results in a "not found" status, low-confidence match, or technical failure costs exactly $0, making it the most cost-effective solution for building autonomous SDR swarms that perform thousands of automated lookups.

Get Started with a Free API Key

Ready to equip your AI agents with live B2B intelligence? Visit lead-enrichment-mcp.agent-infra.workers.dev to generate your API key and access the free tier instantly.

Top comments (0)