Quickstart

This guide walks you through your first HumanConnection search: finding a person and discovering their social profiles.

1. Install the toolkit

$
npm install @humanconnection/toolkit

2. Search for a person

The core tool is search_humans. It takes a name (required) and optional parameters like company_name, company_domain, and location to help identify the right person.

1
import { search_humans } from '@humanconnection/toolkit';
2
3
const results = await search_humans({
4
  name: "Brian Chesky",
5
  company_name: "Airbnb"
6
});
7
8
console.log(JSON.stringify(results, null, 2));

3. Read the results

search_humans returns an array of Human objects. Each contains the person’s name, bio, location, social profiles, and a confidence score. When provide_email is set, results may also include the person’s email and company domain.

1
[
2
  {
3
    "name": "Brian Chesky",
4
    "company_domain": "airbnb.com",
5
    "bio": "Co-Founder & CEO at Airbnb.",
6
    "location": "San Francisco, CA",
7
    "profiles": {
8
      "linkedin_profile_url": "https://linkedin.com/in/brianchesky",
9
      "x_profile_url": "https://x.com/bchesky",
10
      "instagram_profile_url": "https://instagram.com/bchesky"
11
    },
12
    "confidence": 0.85
13
  }
14
]

Understanding the response

How confidence is calculated

A confidence of 0.85 means LinkedIn, X, and Instagram were all found and verified.

4. Try different searches

Use company_name, company_domain, and location to disambiguate people. You can also use people search mode with location and title_filters to find professionals matching specific criteria.

1
// Search by company name
2
await search_humans({ name: "Sara Blakely", company_name: "Spanx" });
3
4
// Search by company domain
5
await search_humans({ name: "David Solomon", company_domain: "goldmansachs.com" });
6
7
// Search by company and location
8
await search_humans({ name: "Melanie Perkins", company_name: "Canva", location: "Sydney" });
9
10
// Search with just a name (works for well-known people)
11
await search_humans({ name: "Elon Musk" });
12
13
// People search mode: find professionals by location and title
14
await search_humans({ location: "Austin", title_filters: "sales AND SaaS" });

5. Use via MCP

The fastest way to give an AI agent access to search_humans is through the remote MCP server.

Claude Desktop

Add to your claude_desktop_config.json:

1
{
2
  "mcpServers": {
3
    "humanconnection": {
4
      "type": "streamable-http",
5
      "url": "https://mcp.humanconnection.sh/mcp"
6
    }
7
  }
8
}

Cursor

Add to .cursor/mcp.json:

1
{
2
  "mcpServers": {
3
    "humanconnection": {
4
      "type": "streamable-http",
5
      "url": "https://mcp.humanconnection.sh/mcp"
6
    }
7
  }
8
}

Once configured, your agent can call search_humans directly from conversation.

Next steps