MCP Server

The Dex MCP Server lets you manage your personal CRM directly from any AI client that supports the Model Context Protocol. Search contacts, log interactions, set reminders, and organize your network — all through natural conversation with your AI assistant.

Claude Code

Claude Desktop

Cursor

VS Code

Gemini CLI

Zed

Also compatible with Codex, Goose, OpenCode, and any MCP-capable client.

Quick Setup

The fastest way to connect is through the hosted MCP server at https://mcp.getdex.com/mcp. No local installation required.

Automatic (Recommended)
Manual Configuration

If you have Node.js installed, run:

npx -y add-mcp "https://mcp.getdex.com/mcp" -y

This auto-detects all supported AI clients on your machine and configures them. Restart your client(s) after running the command. You’ll be prompted to authenticate via browser on first use.

Add the following to your AI client’s MCP configuration file:

{
  "mcpServers": {
    "dex": {
      "type": "http",
      "url": "https://mcp.getdex.com/mcp"
    }
  }
}

Claude Code

Run:

claude mcp add dex --transport http https://mcp.getdex.com/mcp

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).

Cursor

Edit .cursor/mcp.json in your project root or ~/.cursor/mcp.json for global configuration.

VS Code (Copilot)

Edit .vscode/mcp.json in your project root or configure through VS Code settings.

Restart your AI client after configuration. Authentication happens automatically via browser on first connection.

Available Tools

Once connected, your AI assistant gains access to 11 tools for managing your Dex CRM:

Contact Management

Tool	Description
`dex_search_contacts`	Search contacts by name, email, company, or keyword
`dex_get_contact`	Get full contact details with optional interaction history
`dex_create_contact`	Create a new contact
`dex_update_contact`	Update existing contact details
`dex_delete_contacts`	Bulk delete contacts
`dex_merge_contacts`	Merge duplicate contacts

Organization & Timeline

Tool	Description
`dex_manage_tags`	Create, assign, and manage tags across contacts
`dex_manage_groups`	Organize contacts into named collections
`dex_manage_notes`	Log meetings, calls, and interactions on contact timelines
`dex_manage_reminders`	Set follow-up reminders with optional recurrence
`dex_manage_custom_fields`	Define and populate custom data fields

What You Can Do

Here are some examples of what you can ask your AI assistant once connected:

Find Contacts

“Look up Sarah Chen” or “Who do I know at Acme Corp?”

Log Interactions

“Log that I had coffee with Jake today and we discussed the partnership”

Set Reminders

“Remind me to follow up with Maria next Tuesday”

Prep for Meetings

“I have a meeting with David tomorrow — what should I know?”

Organize Network

“Tag everyone from the YC batch as ‘YC W26’”

Clean Up Duplicates

“Find and merge duplicate contacts”

Authentication

The MCP server supports two authentication methods:

Browser OAuth (Default)

On first connection, your AI client will open a browser window for you to log into your Dex account. Credentials are cached for subsequent sessions.

API Key (Headless / CI)

For environments without a browser, use an API key:

Generate an API key

Go to Dex Settings > Integrations.

Set the environment variable

export DEX_API_KEY=dex_your_key_here

API keys require a Dex Professional plan.

Data Model

The MCP server gives your AI assistant access to the full Dex contact graph: Contact ├── Basic Info (name, email, phone, company, job title, birthday) ├── Social Profiles (LinkedIn, Twitter/X, website) ├── Description (rich text notes about the person) ├── Tags (flat labels: “Investor”, “College Friend”) ├── Groups (collections with emoji: ”🏢 Acme Team”) ├── Custom Fields (user-defined: text, dropdown, date) ├── Notes / Timeline (meetings, calls, coffees, general notes) ├── Reminders (follow-ups with optional recurrence) └── Status (starred, archived)

Pagination

All list operations support cursor-based pagination:

limit controls page size (default: 10)
has_more indicates additional results
next_cursor fetches the next page

Response Limits

Responses are capped at 25,000 characters. Use pagination with smaller page sizes for large result sets.

Troubleshooting

Tools not appearing in my AI client

Restart your client after configuring the MCP server. Verify the configuration file is valid JSON and the URL is correct: https://mcp.getdex.com/mcp

Authentication fails or browser doesn't open

Ensure you’re using a supported AI client. For headless environments, use the API key method instead of browser OAuth.

'Professional subscription required' error

Some features require a Dex Professional plan. Check your subscription status in Dex Settings.

Responses are truncated

Use pagination with a smaller limit parameter. Responses exceeding 25,000 characters are automatically truncated.

🚩 Starting with Dex

🤖 AI Features

🛠️ Dex Workflows

⚙️ Integrations

🔄 Sync Features

💰 Sharing Dex

Claude Code

Claude Desktop

Cursor

VS Code

Gemini CLI

Zed

Quick Setup

Available Tools

Contact Management

Organization & Timeline

What You Can Do

Find Contacts

Log Interactions

Set Reminders

Prep for Meetings

Organize Network

Clean Up Duplicates

Authentication

Browser OAuth (Default)

API Key (Headless / CI)

Data Model

Response Limits

Troubleshooting

🚩 Starting with Dex

🤖 AI Features

🛠️ Dex Workflows

⚙️ Integrations

🔄 Sync Features

💰 Sharing Dex

Claude Code

Claude Desktop

Cursor

VS Code

Gemini CLI

Zed

​Quick Setup

​Available Tools

​Contact Management

​Organization & Timeline

​What You Can Do

Find Contacts

Log Interactions

Set Reminders

Prep for Meetings

Organize Network

Clean Up Duplicates

​Authentication

​Browser OAuth (Default)

​API Key (Headless / CI)

​Data Model

​Pagination

​Response Limits

​Troubleshooting

Quick Setup

Available Tools

Contact Management

Organization & Timeline

What You Can Do

Authentication

Browser OAuth (Default)

API Key (Headless / CI)

Data Model

Pagination

Response Limits

Troubleshooting