Introduction
Klozeo is an API-first lead database built for teams that want to manage, score, and query their business leads programmatically — including through AI assistants like Claude and Cursor.
If you’ve ever found yourself juggling lead data across spreadsheets, CRMs, and custom scripts, Klozeo gives you a single, structured place to store that data and a powerful API to work with it.
What problem does Klozeo solve?
Section titled “What problem does Klozeo solve?”Most tools that store lead data were designed for humans — forms, tables, manual imports. They’re hard to automate, hard to query precisely, and impossible to connect to AI workflows without significant glue code.
Klozeo is designed for a different world:
- Your pipeline pushes leads via API — from scrapers, webhooks, or integrations
- Your rules score them automatically — based on criteria you define in plain expressions
- Your AI assistant queries them in natural language — using the built-in MCP integration
You stay in control. Klozeo handles the storage, deduplication, scoring, and exposure.
Core concepts
Section titled “Core concepts”Before diving into the API, here are the five concepts you’ll encounter everywhere:
A lead is any business or person you want to track. Each lead has standard fields (name, email, phone, website, city, country…) plus any custom fields you add via attributes.
Leads are identified by a prefixed ID: cl_<uuid>.
Attributes
Section titled “Attributes”Attributes are custom fields you attach to a lead. They let you extend the default schema without any migrations — add cuisine, employee_count, last_meeting_date, or any field your workflow needs.
Scoring rules
Section titled “Scoring rules”A scoring rule is an expression that assigns a numeric weight to a lead property. Klozeo evaluates all your rules and computes a score from 0 to 100 for each lead.
Example: “Add 20 points if the lead has a website. Add 30 more if their city is Paris.”
Deduplication
Section titled “Deduplication”When you push a new lead, Klozeo checks automatically whether it already exists — by email, source ID, phone, or name+city — and merges it rather than creating a duplicate. You get clean data without extra logic on your side.
MCP integration
Section titled “MCP integration”MCP (Model Context Protocol) lets AI clients like Claude or Cursor connect directly to Klozeo and query your leads in natural language. Instead of writing API calls, you ask: “Show me the top 10 leads in Berlin with a score above 70.”
How the documentation is organized
Section titled “How the documentation is organized”| Section | What you’ll find |
|---|---|
| Getting Started | Authentication, your first API call, rate limits |
| API Reference | Every endpoint, parameter, and response format |
| Guides | Deep dives: filtering, pagination, deduplication, batch ops, scoring |
| MCP Integration | How to connect Claude, Cursor, Cline, and other AI clients |
| SDKs | Official TypeScript, Python, Go, and Rust client libraries |
| Resources | FAQ and changelog |
Where to start
Section titled “Where to start”I want to make my first API call → Authentication then Your first request
I want to connect Claude or Cursor → MCP Overview
I want to understand how scoring works → Lead Scoring guide
I want to push leads in bulk → Batch Operations
I have a specific question → FAQ
API at a glance
Section titled “API at a glance”Base URL: https://api.klozeo.com/api/v1Auth: X-API-Key: <your-key>Format: JSONAll endpoints follow REST conventions. Errors return a consistent { "error": "code", "message": "..." } shape. Lead IDs are prefixed (cl_), note IDs are prefixed (note_), everything else is a plain UUID.