Supercharging Cursor and VS Code with a Live B2B Lead Enrichment MCP Server

The most efficient way to give AI agents native access to live B2B firmographics and technographic data without custom middleware is by deploying a Model Context Protocol (MCP) server. By integrating a dedicated B2B lead enrichment MCP server into development environments like Cursor or VS Code, you provide your LLM with real-time company profiles and intent signals directly through a standardized JSON-RPC interface.

Core Features

This MCP implementation bridges the gap between static LLM knowledge and the dynamic B2B landscape. It uses a strictly typed schema to ensure that AI agents can query and process lead data with high precision.

• Deep Firmographics: Access real-time data including company size, exact headcount, verified industry classifications, and headquarters location.
• Technographic Intelligence: Identify the specific software stack used by a target organization (e.g., CRM choices, cloud providers, and analytics tools).
• Intent Signal Processing: Retrieve active buying signals and recent organizational changes that indicate a high propensity to convert.
• Hallucination Prevention: All tool parameters are defined using strict Zod annotations, forcing the LLM to adhere to valid search strings and domain formats, virtually eliminating parameter-based hallucinations.

IDE Configuration

To enable these tools in your AI-powered IDE (like Cursor or Claude Desktop), add the following configuration to your settings file. This points your environment to the hosted MCP endpoint.

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

Standardized JSON-RPC Response

When your agent invokes the enrich_lead tool, the MCP server returns a structured payload. This clean data structure allows the AI to immediately reason about the lead without parsing messy HTML or unformatted text.

{
  "jsonrpc": "2.0",
  "id": "1",
  "result": {
    "companyName": "Acme Corp",
    "industry": "Enterprise SaaS",
    "technographics": ["Salesforce", "AWS", "Segment", "React"],
    "intentSignals": [
      { "type": "funding_round", "detail": "Series C - $50M", "date": "2023-10-15" },
      { "type": "hiring_spree", "detail": "15+ openings in DevOps", "confidence": 0.92 }
    ],
    "confidenceScore": 0.89,
    "metadata": { "lastUpdated": "2023-11-01T10:00:00Z" }
  }
}

Risk-Free Metered Billing

Modern AI workflows require cost predictability. This MCP server utilizes a performance-based billing model. Users are only debited for "High Confidence" results. If the enrichment engine returns a result with a Confidence Score < 0.6 (indicating low-quality or unverifiable data), the query is processed at zero cost. This ensures your autonomous SDR agents or research tools only consume credits when the data is actionable and accurate.

Get Started with a Free API Key

Ready to bridge the gap between your LLM and live B2B intelligence? You can get up and running in less than two minutes by grabbing a free tier API key.

Visit lead-enrichment-mcp.agent-infra.workers.dev to generate your credentials and start enriching your AI workflows today.