You're already collecting name, email, and phone. Here's what those three fields are actually worth.

# Works with Claude, Cursor, Antigravity — any IDE/agent stack.
# Generate a token, hand this to your agent, it self-installs in seconds.

> "Pre-qualify the new leads in my CRM and route the SQLs to the closer calendar."

# Your agent calls the LeadFi MCP tool:
leadfi.prequalify({
  name:  "Jordan Marsh",
  email: "jordan@northway.co",
  phone: "+15125550142"
})
# <- { classification: "SQL", tier: "Tier 2", route: "calendar" }

# Pre-qualify a lead you already captured.
curl -X POST https://api.leadfi.ai/v1/prequalify \
  -H "Authorization: Bearer $LEADFI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name":  "Jordan Marsh",
    "email": "jordan@northway.co",
    "phone": "+15125550142",
    "consent_ref": "consent_9f12ab"
  }'

import { LeadFi } from "@leadfi/sdk";

const leadfi = new LeadFi(process.env.LEADFI_API_KEY);

const result = await leadfi.prequalify({
  name:  "Jordan Marsh",
  email: "jordan@northway.co",
  phone: "+15125550142",
  consentRef: "consent_9f12ab",
});

// -> { classification, tier, route, consent_ref }
routeLead(result.classification, result.route);

from leadfi import LeadFi

leadfi = LeadFi(api_key=os.environ["LEADFI_API_KEY"])

result = leadfi.prequalify(
    name="Jordan Marsh",
    email="jordan@northway.co",
    phone="+15125550142",
    consent_ref="consent_9f12ab",
)

# result.classification -> "SQL" | "NQL"
route_lead(result.classification, result.route)