Gong Interaction Data Bridge

This Model Context Protocol (MCP) server acts as an abstraction layer to access the Gong service API, facilitating the extraction of recorded customer interactions and their associated transcriptions. It enables large language models like Claude to systematically engage with proprietary Gong data.

Core Capabilities

• Query a manifest of Gong interactions, optionally constrained by specified time windows.
• Retrieve exhaustive textual renditions (transcripts) for selected interaction identifiers.
• Secure access is managed via mandatory Gong API credentials.
• Provides a standardized MCP endpoint structure for seamless integration into AI workflows.

Setup Requirements

• Node.js environment, version 18 or newer.
• Docker utility (recommended for deployment via containerization).
• Valid Gong API access credentials (a key and a secret).

Deployment Instructions

Local Execution

1. Obtain a local copy of the source repository.
2. Install required software dependencies: bash npm install
3. Compile the source code: bash npm run build

Containerized Deployment

Build the deployment image:

docker build -t gong-mcp .

Configuration for Claude Integration

1. Access the settings panel within the Claude application.
2. Navigate to the section dedicated to MCP Servers.
3. Establish a new server configuration using the following blueprint:

{
  "command": "docker",
  "args": [
    "run",
    "-it",
    "--rm",
    "gong-mcp"
  ],
  "env": {
    "GONG_ACCESS_KEY": "your_actual_access_key",
    "GONG_ACCESS_SECRET": "your_actual_secret_token"
  }
}

1. Substitute the placeholder values above with your verifiable Gong API credentials, typically sourced from a configuration file (e.g., .env).

Exposed Functionality (Tools)

Query Interaction Log

Fetches a directory of Gong interactions, permitting temporal boundaries to refine the results set.

{
  "name": "list_calls",
  "description": "Obtain a manifest of Gong interactions, optionally filtered by a date range. Returns metadata such as interaction ID, subject line, initiation/termination timestamps, participant roster, and total duration.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "fromDateTime": {
        "type": "string",
        "description": "The starting temporal marker in ISO 8601 format (e.g., 2024-03-01T00:00:00Z)"
      },
      "toDateTime": {
        "type": "string",
        "description": "The ending temporal marker in ISO 8601 format (e.g., 2024-03-31T23:59:59Z)"
      }
    }
  }
}

Fetch Detailed Dialogue Content

Retrieves the complete textual conversation record associated with specific interaction identifiers.

{
  "name": "retrieve_transcripts",
  "description": "Acquire the comprehensive textual record for designated interaction IDs. Output includes speaker segmentation, identified subjects, and time-stamped utterance segments.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "callIds": {
        "type": "array",
        "items": { "type": "string" },
        "description": "A list of unique Gong interaction identifiers for which transcripts are requested"
      }
    },
    "required": ["callIds"]
  }
}

Licensing

This project is governed by the MIT License; refer to the LICENSE documentation file for specifics.

Contribution Guide

1. Fork the primary repository.
2. Establish a new feature branch (git checkout -b feature/my-enhancement).
3. Commit the modifications (git commit -m 'Implement feature X enhancement').
4. Push the branch to the upstream (git push origin feature/my-enhancement).
5. Submit a Pull Request for review.