---
updatedAt: 2026-06-18T15:43:06.000Z
agentTools:
  projectIndex: https://docs.fullcontact.com/llms.txt
---

#  MCP Server Overview

The FullContact **Model Context Protocol (MCP) Server** brings FullContact's identity and enrichment data directly into AI agents and AI-native developer tools. With a one-time configuration step, your existing FullContact API key powers an `enrich_person` tool that any MCP-compatible client — Claude Desktop, Claude Code, Cursor, VS Code, ChatGPT, Gemini(CLI only at this time) and others — can call as part of an agentic workflow.

The server is a thin, secure wrapper around FullContact's `person.enrich` API. It reuses the authentication, rate limits, and compliance controls that already govern your account, so you can adopt it without provisioning new credentials or signing new agreements.

## What you can do with it

Drawing from our patented identity graph, the MCP Server empowers your AI agents to:

* **Resolve a person from a single identifier.** Pass an email, phone number, social profile, or name + postal address and receive a unified person profile.
* **Enrich CRM and marketing workflows.** Let an agent watch new signups, leads, or support tickets and append demographic, employment, and social data on the fly.
* **Power conversational research.** Ask your IDE assistant or chatbot, "Who is this person, and where do they work?" and get a structured answer grounded in FullContact data.
* **Stay inside your existing controls.** All calls are authenticated with your FullContact API key and counted against your normal rate limits and billing plan.

## How it works

<Image src="https://files.readme.io/ef4fbe9805b0ae72b70e5617d4193525425ee09d077bfcdd60c63e33e7f5ca16-Screenshot_2026-04-29_at_2.22.38_PM.png" align="center" />

1. The AI client discovers the server's tool catalog and JSON Schema for `enrich_person`.
2. When the agent decides to enrich a person, it sends an MCP request through the configured client.
3. The MCP Server validates your API key, applies your account's rate limits, and forwards the request to `person.enrich`.
4. The response is normalized into a compact, agent-friendly JSON payload and returned to the client.

<br />

## Available tools

| Tool            | Purpose                                                                    | Underlying API                                                                    |
| --------------- | -------------------------------------------------------------------------- | --------------------------------------------------------------------------------- |
| `enrich_person` | Resolve and enrich a person from email, phone, social, or name + location. | [`POST /v3/person.enrich`](https://docs.fullcontact.com/docs/multi-field-request) |

Additional tools (company enrichment, Resolve, Acumen) will follow in future releases.

<br />

## Endpoint

The hosted MCP Server is available at:

```http
https://api.fullcontact.com/v3/mcp
```

Authenticate with your existing FullContact API key — see [Authentication](./03-mcp-authentication.md). There is no infrastructure for you to operate.

## Requirements

* An active FullContact account with API access.
* A FullContact API key with `person.enrich` enabled. See [Generate an API Key](https://docs.fullcontact.com/docs/generate-an-api-key).
* An MCP-compatible client (Claude Desktop, Claude Code, Cursor, VS Code with the MCP extension, ChatGPT with Connectors, or any other MCP-aware tool).

## Next steps

* [Quickstart](https://docs.fullcontact.com/docs/quickstart) — Connect the MCP Server to your AI client in under five minutes.
* [Authentication](https://docs.fullcontact.com/docs/authentication) — Where to obtain and how to pass your API key.
* [Tool Reference: `enrich_person`](https://docs.fullcontact.com/docs/tool-reference-enrich_person) — Inputs, outputs, and example payloads.

***

*Updated May 2026*