Gong-Mcp-Server

Tailored-AI-Hub/Gong-Mcp-Server

3.2

If you are the rightful owner of Gong-Mcp-Server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The Gong MCP Server integrates Claude with Gong, enabling natural language interactions with Gong call data and analytics.

Tools
8
Resources
0
Prompts
0

Gong MCP Server

An MCP (Model Context Protocol) server implementation that integrates Claude with Gong, enabling natural language interactions with your Gong call data and analytics. This server allows Claude to query, analyze, and generate insights from your sales conversations and team performance using everyday language.

Features

  • Call Management: List and retrieve detailed information about calls within date ranges
  • Transcript Analysis: Get full call transcripts for content analysis and insights
  • User Management: Access user information and team structure
  • Activity Analytics: Generate comprehensive activity statistics and performance metrics
  • Period-based Analysis: Analyze data by day, week, month, quarter, or year
  • Scorecard Generation: Create performance scorecards with metrics and insights
  • Interaction Patterns: Analyze call interaction patterns and behaviors
  • Flexible Filtering: Filter data by users, call types, directions, and time periods

Installation

npm install -g @tailoredai/gong-mcp-server

Or run on-demand without global install:

npx -y @tailoredai/gong-mcp-server

Tools

gong_list_calls

List recent calls. No inputs.

  • Response: pagination summary (may be omitted) then entries like:
    • Call ID, Title, Direction, Scheduled Time, Started Time, Duration, Call Purpose, Language

gong_get_call

Get a specific call by ID.

  • Input: { "callId": "2167868958109749118" }
  • Response: one formatted call entry with the same fields as above

gong_get_transcript

Get transcripts for one or more calls.

  • Input: { "filter": { "callIds": ["379333695432645797", "2167868958109749118"] } }
  • Response: for each call:
    • Transcript for Call ID: ...
    • Speaker ID: ... lines
    • Sentences with timestamps like [mm:ss - mm:ss] text

gong_list_users

List users with optional pagination.

  • Optional inputs: pageNumber, pageSize, cursor, includeInactive
  • Response: pagination summary then entries:
    • User ID, Name, Email, Title, Manager ID, Active

gong_get_user

Get a specific user by ID.

  • Input: { "userId": "1587172352477568464" }
  • Response: User Details: block with ID, Name, Email, Title, Manager ID, Active

gong_get_activity_aggregate

Get user-level activity totals in a date range.

  • Input: { "filter": { "fromDate": "YYYY-MM-DD", "toDate": "YYYY-MM-DD", "__userIds": [optional] } }
  • Response: header Activity Statistics (from to), then User Totals blocks per user with fields like callsAsHost, callsAttended, feedback/listening/share stats

gong_get_activity_scorecards

Fetch activity scorecards.

  • Input: { "aggregationPeriod": "QUARTER", "filter": { "fromDate": "YYYY-MM-DD", "toDate": "YYYY-MM-DD", "__userIds": [optional] } }
  • Response: scorecard blocks with period, score, and Answers: list (question/score/overall)

gong_get_interaction_stats

Per-person interaction metrics within a date range.

  • Input: { "filter": { "fromDate": "YYYY-MM-DD", "toDate": "YYYY-MM-DD", "__userIds": [optional] } }
  • Response: for each person: User ID, User Email, then metric lines like - Talk Ratio: 65

Setup

Gong Authentication

You need to set up your Gong API credentials:

  1. Get your Gong Access Key and Access Key Secret from your Gong account
  2. Set environment variables for authentication

Usage with Claude Desktop

Add to your claude_desktop_config.json (npx):

{
  "mcpServers": {
    "gong": {
      "command": "npx",
      "args": ["-y", "@tailoredai/gong-mcp-server"],
      "env": {
        "GONG_ACCESS_KEY": "your_access_key",
        "GONG_ACCESS_KEY_SECRET": "your_access_key_secret",
        "GONG_BASE_URL": "https://api.gong.io/v2"
      }
    }
  }
}

Or using the globally installed binary (no npx):

{
  "mcpServers": {
    "gong": {
      "command": "gong-connector",
      "env": {
        "GONG_ACCESS_KEY": "your_access_key",
        "GONG_ACCESS_KEY_SECRET": "your_access_key_secret",
        "GONG_BASE_URL": "https://api.gong.io/v2"
      }
    }
  }
}

Example Usage

Call Analysis

"List all calls from last month"
"Get transcript for the most recent sales call"
"Show me calls longer than 30 minutes from this week"

User Performance

"Get activity stats for the sales team this quarter"
"Show me the top performers by call volume"
"Generate a scorecard for user 1234567890"

Team Analytics

"Compare team performance by month for 2022"
"Analyze call patterns by department"
"Get daily activity breakdown for the last 30 days"

Transcript Analysis

"Get transcripts for all calls with prospects this week"
"Analyze conversation patterns in discovery calls"
"Review call quality metrics for the team"

Development

Building from source

# Clone the repository
git clone https://github.com/Tailored-AI-Hub/Gong-Mcp-Server

# Navigate to directory
cd Gong-Mcp-Server

# Install dependencies
npm install

# Build the project
npm run build

Environment Variables

  • GONG_ACCESS_KEY: Your Gong API access key
  • GONG_ACCESS_KEY_SECRET: Your Gong API access key secret
  • GONG_BASE_URL: Gong Base URL (default: https://api.gong.io/v2)

API Endpoints Covered

This MCP server covers the following Gong API endpoints:

  • GET /v2/calls - List calls with date filtering
  • GET /v2/calls/{id} - Get specific call details
  • POST /v2/calls/transcript - Get call transcripts
  • GET /v2/users - List organization users
  • GET /v2/users/{id} - Get specific user details
  • POST /v2/stats/activity/aggregate - Aggregate activity statistics
  • POST /v2/stats/activity/aggregate-by-period - Period-based aggregation
  • POST /v2/stats/activity/day-by-day - Daily activity breakdown
  • POST /v2/stats/activity/scorecards - Performance scorecards
  • POST /v2/stats/interaction - Interaction pattern analysis

Contributing

Contributions are welcome! Please open an issue to discuss major changes and submit a PR.

  1. Fork the repo and create a feature branch.
  2. Run tests with npm test and ensure they pass.
  3. Follow conventional commits if possible.

License

This project is licensed under the MIT License - see the file for details.

Issues and Support

If you encounter any issues or need support, please file an issue on the GitHub repository.