Intercom

MCP Server for Intercom

What is MCP Server for Intercom?

MCP Server for Intercom is an MCP-compliant server designed to enable AI assistants, such as Claude Desktop and Cline, to access and analyze customer support tickets from Intercom. It acts as a bridge between Intercom's customer support data and AI tools, facilitating efficient data retrieval and analysis.

How to use MCP Server for Intercom?

To use MCP Server for Intercom, you need Node.js 18.0.0 or higher and an Intercom account with API access. You can install it globally via npm: npm install -g mcp-server-for-intercom, then set your Intercom API token: export INTERCOM_ACCESS_TOKEN="your_token_here", and run the server: intercom-mcp. Alternatively, you can use Docker by building the image and running the container with your API token and port mappings. For integration with Claude Desktop, you can configure it in your claude_desktop_config.json file.

Key Features of MCP Server for Intercom

• Advanced Search and Filtering: Search conversations and tickets with advanced filtering options, including by customer, status, date range, and keywords.
• Email Content Search: Ability to search by email content even when no contact exists in Intercom.
• Efficient Server-Side Filtering: Utilizes Intercom's search API for efficient server-side filtering.
• Seamless AI Integration: Designed for seamless integration with MCP-compliant AI assistants.
• Available MCP Tools: Provides specific tools like list_conversations (retrieves conversations within a date range with content filtering), search_conversations_by_customer (finds conversations for a specific customer), search_tickets_by_status (retrieves tickets by their status), and search_tickets_by_customer (finds tickets associated with a specific customer).

Use Cases of MCP Server for Intercom

• AI-Powered Customer Support Analysis: AI assistants can access and analyze Intercom support tickets to gain insights into customer issues, trends, and sentiment.
• Automated Reporting: Generate automated reports on customer conversations and ticket statuses.
• Enhanced Customer Service: Improve customer service by providing AI tools with direct access to support data for faster resolution and personalized responses.
• Developer Integration: Developers can leverage the provided MCP tools to build custom applications and integrations that interact with Intercom data.

FAQ from the Project

• What are the prerequisites for installation? Node.js 18.0.0 or higher, an Intercom account with API access, and your Intercom API token.
• How do I set my Intercom API token? You can set it as an environment variable: export INTERCOM_ACCESS_TOKEN="your_token_here" or pass it during Docker container execution.
• What is the difference between the default Docker version and the standard version? The default Docker configuration is optimized for Glama compatibility and includes specific dependencies, while the standard version is a lighter version without Glama-specific dependencies.
• What are the limitations for list_conversations? The date range must not exceed 7 days.
• Is this project affiliated with Intercom Inc.? No, this project is an independent integration and is not affiliated with, officially connected to, or endorsed by Intercom Inc.

MCP Server for Intercom

An MCP-compliant server that enables AI assistants to access and analyze customer support data from Intercom.

Features

• Search conversations and tickets with advanced filtering
• Filter by customer, status, date range, and keywords
• Search by email content even when no contact exists
• Efficient server-side filtering via Intercom's search API
• Seamless integration with MCP-compliant AI assistants

Installation

Prerequisites

• Node.js 18.0.0 or higher
• An Intercom account with API access
• Your Intercom API token (available in your Intercom account settings)

Quick Setup

Using NPM

# Install the package globally
npm install -g mcp-server-for-intercom

# Set your Intercom API token
export INTERCOM_ACCESS_TOKEN="your_token_here"

# Run the server
intercom-mcp

Using Docker

The default Docker configuration is optimized for Glama compatibility:

# Start Docker (if not already running)
# On Windows: Start Docker Desktop application
# On Linux: sudo systemctl start docker

# Build the image
docker build -t mcp-intercom .

# Run the container with your API token and port mappings
docker run --rm -it -p 3000:3000 -p 8080:8080 -e INTERCOM_ACCESS_TOKEN="your_token_here" mcp-intercom:latest

Validation Steps:

# Test the server status
curl -v http://localhost:8080/.well-known/glama.json
# Test the MCP endpoint
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1,"method":"mcp.capabilities"}' http://localhost:3000

Alternative Standard Version

If you prefer a lighter version without Glama-specific dependencies:

# Build the standard image
docker build -t mcp-intercom-standard -f Dockerfile.standard .

# Run the standard container
docker run --rm -it -p 3000:3000 -p 8080:8080 -e INTERCOM_ACCESS_TOKEN="your_token_here" mcp-intercom-standard:latest

The default version includes specific dependencies and configurations required for integration with the Glama platform, while the standard version is more lightweight.

Available MCP Tools

1. list_conversations

Retrieves all conversations within a date range with content filtering.

Parameters:

• startDate (DD/MM/YYYY) – Start date (required)
• endDate (DD/MM/YYYY) – End date (required)
• keyword (string) – Filter to include conversations with this text
• exclude (string) – Filter to exclude conversations with this text

Notes:

• Date range must not exceed 7 days
• Uses efficient server-side filtering via Intercom's search API

Example:

{
  "startDate": "15/01/2025",
  "endDate": "21/01/2025",
  "keyword": "billing"
}

2. search_conversations_by_customer

Finds conversations for a specific customer.

Parameters:

• customerIdentifier (string) – Customer email or Intercom ID (required)
• startDate (DD/MM/YYYY) – Optional start date
• endDate (DD/MM/YYYY) – Optional end date
• keywords (array) – Optional keywords to filter by content

Notes:

• Can find conversations by email content even if no contact exists
• Resolves emails to contact IDs for efficient searching

Example:

{
  "customerIdentifier": "customer@example.com",
  "startDate": "15/01/2025",
  "endDate": "21/01/2025",
  "keywords": ["billing", "refund"]
}

3. search_tickets_by_status

Retrieves tickets by their status.

Parameters:

• status (string) – "open", "pending", or "resolved" (required)
• startDate (DD/MM/YYYY) – Optional start date
• endDate (DD/MM/YYYY) – Optional end date

Example:

{
  "status": "open",
  "startDate": "15/01/2025",
  "endDate": "21/01/2025"
}

4. search_tickets_by_customer

Finds tickets associated with a specific customer.

Parameters:

• customerIdentifier (string) – Customer email or Intercom ID (required)
• startDate (DD/MM/YYYY) – Optional start date
• endDate (DD/MM/YYYY) – Optional end date

Example:

{
  "customerIdentifier": "customer@example.com",
  "startDate": "15/01/2025",
  "endDate": "21/01/2025"
}

Configuration with Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "intercom-mcp": {
      "command": "intercom-mcp",
      "args": [],
      "env": {
        "INTERCOM_ACCESS_TOKEN": "your_intercom_api_token"
      }
    }
  }
}

Implementation Notes

For detailed technical information about how this server integrates with Intercom's API, see src/services/INTERCOM_API_NOTES.md. This document explains our parameter mapping, Intercom endpoint usage, and implementation details for developers.

Development

# Clone and install dependencies
git clone https://github.com/raoulbia-ai/mcp-server-for-intercom.git
cd mcp-server-for-intercom
npm install

# Build and run for development
npm run build
npm run dev

# Run tests
npm test

Disclaimer

This project is an independent integration and is not affiliated with, officially connected to, or endorsed by Intercom Inc. "Intercom" is a registered trademark of Intercom Inc.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.