mcp -for-intercom

This MCP server provides an interface for AI assistants to access and analyze customer support data stored in Intercom. It exposes tools to search conversations and tickets with advanced filtering, including date ranges, keywords, statuses, and customer identifiers. The available tools are designed to let a client agent or an AI workflow quickly retrieve relevant interaction history, resolve emails to Intercom contact IDs, and filter results server-side via Intercom's APIs for efficiency. To use it, deploy the server (via npm, Docker, or your preferred method) and then call the MCP endpoints or tools exposed by the server as part of your AI assistant’s planning and reasoning steps.

Once running, you can leverage the following MCP tools:

• list_conversations: fetch conversations within a date range with optional keyword filters.
• search_conversations_by_customer: locate conversations tied to a specific customer (by email or Intercom ID), with optional date and keyword filters.
• search_tickets_by_status: retrieve tickets by status (open, pending, resolved) with optional date filtering.
• search_tickets_by_customer: find tickets associated with a given customer (email or Intercom ID) with optional date filtering.

These tools enable comprehensive data retrieval while minimizing client-side processing. Use the provided parameters to tailor requests (e.g., date ranges limited to 7 days for list_conversations) and rely on the server-side Intercom API calls for performance and accuracy.

Prerequisites:

• Node.js 18.0.0 or higher
• An Intercom account with API access
• Your Intercom API token (INTERCOM_ACCESS_TOKEN)

Installation options:

Option 1: Quick npm setup (global CLI)

1. Install the MCP server globally:

npm install -g mcp-server-for-intercom

2. Set your Intercom API token:

export INTERCOM_ACCESS_TOKEN="your_token_here"

3. Run the server (CLI command provided by the package):

intercom-mcp

Option 2: Docker (default Glama-compatible configuration)

1. Build the image:

docker build -t mcp-intercom .

2. 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

Option 2 alternative: Standard version (lighter, no Glama dependencies)

1. Build the standard image:

docker build -t mcp-intercom-standard -f Dockerfile.standard .

2. 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

Development/Testing (from source):

# 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

Prerequisites note: Ensure your environment has Docker installed if you plan to run the Docker-based setup.

Tips and considerations:

• The Intercom API token must be kept secret; avoid committing it to version control.
• The MCP endpoints expect specific JSON-RPC shaped payloads; use the provided examples to format requests.
• For Glama integration scenarios, the default image includes dependencies and configurations geared toward Glama compatibility; if you don’t need Glama, use the standard docker variant.
• If you encounter authentication errors, verify that INTERCOM_ACCESS_TOKEN is valid and that Intercom API access is enabled for the token.
• The list_conversations tool has a 7-day maximum range per request; plan pagination or multiple calls for longer ranges.
• When deploying behind a proxy or firewall, ensure the exposed ports (commonly 3000 and 8080) are accessible and properly mapped.