Anthropic's Claude Code just got a significant upgrade for B2B sales and revenue teams. The @agent-infra/mcp-server-lead-enrichment package integrates directly into Claude Code, enabling real-time access to firmographic data, technographic insights, and buyer intent signals without leaving the AI coding environment.

What This Integration Delivers

The MCP server brings three core data categories into Claude Code workflows. Firmographics cover company size, industry, revenue range, and organizational structure—essential context for any outbound or account-based strategy. Technographics reveal what tools and platforms a target company uses, including marketing automation, CRM, hosting providers, and development frameworks. Intent signals surface buying behavior indicators from content consumption patterns, helping teams prioritize outreach based on actual purchase consideration rather than guesswork.

The Confidence-Based Billing Model

The pricing model is worth noting: you pay only when confidenceScore exceeds 0.6. This threshold acts as a quality gate—if the enrichment engine can't produce reliable data, there's no charge. For high-volume workflows where you're processing dozens or hundreds of leads, this approach eliminates wasted spend on low-quality matches and ensures teams are acting on verified intelligence rather than guesswork dressed up as data.

Why It Matters for AI Agent Workflows

This integration reflects a broader shift toward embedding third-party data directly into AI agent pipelines. Rather than manually exporting enrichment results or switching between tools, Claude Code can now pull live B2B context mid-task—whether that's drafting personalized outreach sequences, preparing account research summaries, or qualifying inbound leads automatically.

Key Takeaways

  • Real-time firmographics, technographics, and intent signals directly in Claude Code workflows
  • Pay-per-quality model: charges only when confidenceScore > 0.6
  • Zero cost for low-confidence matches prevents wasted spend on poor data
  • Enables AI-native sales intelligence without leaving the coding environment

The Bottom Line

This isn't revolutionary—lead enrichment APIs have existed for years—but embedding pay-per-quality B2B data into Claude Code signals that AI agents are becoming genuine revenue workflow tools, not just code assistants. Teams already living in Claude Code will find this a natural fit; everyone else should watch how MCP server integrations evolve as enterprises demand tighter agent-to-data pipelines.