Skip to main content
The peeps MCP server is distributed as an npm package (@peeps-ai/mcp) and can be run with npx, installed globally, or skipped entirely in favor of the peeps-hosted remote server. Pick the option that best fits your environment — all three expose the same tools to your AI assistant.

​
Prerequisites

Before you install, make sure you have the following:
  • Node.js 18 or later — required for the npx and global install options. Run node --version to check.
  • A peeps account — sign up at app.peeps.ai if you don’t have one.
  • A peeps API key — find yours in the peeps dashboard settings.
You can grab your API key from Settings → API Keys inside the peeps dashboard. Keep it somewhere safe — you’ll need it in every installation option below.

​
Installation options

1

Choose your installation method

Pick one of the three options below based on your environment. Most users should start with Option 1 (npx).
2

Set your API key and run the server

3

Verify the connection

Once the server is running (or your client is pointed at the remote URL), ask your AI assistant a question that requires contact data — for example: “Search for engineers at Acme Corp.” A successful response means the MCP server is connected and authenticated.

​
Option details

npx runs the MCP server without a permanent installation. Your MCP client (Claude Desktop, Cursor, etc.) typically handles launching the process for you — you just supply the command and your API key in the client’s config file.
The -y flag skips the confirmation prompt so the server starts immediately. This is the method used in the Claude Desktop and Cursor configuration examples.

​
Option 2: Install globally

If you prefer a persistent, version-pinned installation, install the package globally with npm:
Then start the server with:
To upgrade to a newer version later:

​
Option 3: Use the remote server

Connect directly to the peeps-hosted MCP server without installing anything locally. This option is ideal for cloud-hosted agents and serverless environments.
Authenticate by passing your API key as a Bearer token in the Authorization header:
Refer to your MCP client’s documentation for how to configure custom headers on an SSE transport connection. See the configuration guide for a generic JSON example.
Never hard-code your API key in a script or config file that gets committed to version control. Use environment variables or a secrets manager instead. See the configuration guide for safe patterns.