> ## Documentation Index
> Fetch the complete documentation index at: https://docs.peepsai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# peeps MCP server overview: people and contact data

> The peeps MCP server connects AI assistants to your contact data. Learn what it can do, how it works, and which transport option suits your setup.

The peeps MCP server connects your AI assistant directly to your people and contact data. By implementing the [Model Context Protocol](https://modelcontextprotocol.io), it lets tools like Claude, Cursor, and any other MCP-compatible client search, retrieve, and reason over your contacts and organizations — all through natural language, without writing a single line of integration code.

## What you can do

With the peeps MCP server connected to your AI assistant, you can:

* **Search people** by name, job title, email address, or the organization they belong to
* **Look up a contact's full profile** — including phone, LinkedIn URL, and organizational affiliation — by their ID
* **Search organizations** by name or domain to identify companies and their details
* **Retrieve an organization's members** in a single call when you need to understand a team or account
* **Paginate through your entire contact list** sorted by name, creation date, or last-updated date
* **Filter searches by organization** to scope results to a specific company

## How it works

When your AI assistant needs people data, it follows this path:

```
AI Assistant (Claude, Cursor, etc.)
        │
        │  calls MCP tool (e.g. search_people)
        ▼
  MCP Client Layer
        │
        │  stdio pipe (local) or HTTP/SSE (remote)
        ▼
peeps MCP Server (@peeps-ai/mcp)
        │
        │  authenticated REST request
        ▼
   peeps API (api.peeps.ai)
        │
        │  returns structured JSON
        ▼
  MCP Client Layer
        │
        │  structured tool result
        ▼
AI Assistant (reasons over the data)
```

Your API key authenticates every request from the MCP server to the peeps API. The AI assistant never calls the peeps API directly — the MCP server handles all authentication and data shaping before returning clean, structured results.

## Transport options

The peeps MCP server supports two transport modes. Choose the one that matches where your AI assistant runs.

### stdio — local installation

In stdio mode, you run the MCP server as a local process using `npx`. Your MCP client (Claude Desktop, Cursor, etc.) spawns the process and communicates with it over standard input/output.

**Best for:** Claude Desktop, Cursor, local development environments, and any MCP client that manages its own server processes.

```bash theme={null}
PEEPS_API_KEY=your-api-key npx -y @peeps-ai/mcp
```

### HTTP/SSE — remote server

In HTTP/SSE mode, you connect to the peeps-hosted MCP server at `https://mcp.peeps.ai/sse`. Nothing is installed locally — your MCP client connects over the network using Server-Sent Events.

**Best for:** Cloud-hosted AI agents, serverless environments, and any setup where installing a local process is not practical.

```
https://mcp.peeps.ai/sse
```

Pass your API key as a Bearer token in the `Authorization` header when connecting.

<Tip>
  If you're just getting started, stdio mode with `npx` is the fastest path. You don't need a global install — `npx` downloads and runs the latest version automatically.
</Tip>

## Next steps

<CardGroup cols={2}>
  <Card title="Installation" icon="download" href="/mcp/installation">
    Install the peeps MCP server via npx, npm, or connect to the remote server — no local install required.
  </Card>

  <Card title="Configuration" icon="sliders" href="/mcp/configuration">
    Set your API key and configure Claude Desktop, Cursor, or any MCP-compatible client.
  </Card>

  <Card title="Tools Reference" icon="wrench" href="/mcp/tools">
    Explore every tool the MCP server exposes, with parameters, return fields, and example prompts.
  </Card>

  <Card title="Quickstart" icon="bolt" href="/quickstart">
    Go from zero to your first AI-powered contact search in under five minutes.
  </Card>
</CardGroup>