Gamma MCP Server

A Model Context Protocol (MCP) server that integrates Gamma.app with AI assistants. Create presentations, documents, webpages, and social posts directly from your AI conversations.

Works with: Claude Code, Claude Desktop, OpenCode, GitHub Copilot CLI, Google Gemini CLI, and any other MCP-compatible AI assistant.

Features

• Generate Content: Create professional presentations, documents, webpages, and social posts from text prompts
• Theme Support: Browse and apply visual themes to your content
• Folder Organization: Save generated content to specific folders
• Template Remix: Create variations of existing Gamma templates
• Email Sharing: Share generated content directly via email

Quick Start

1. Clone and Install

git clone https://github.com/Arkava-AI/gamma-mcp-server.git
cd gamma-mcp-server
npm install
npm run build

2. Get Your Gamma API Key

1. Log in to gamma.app
2. Go to Settings > API (or Settings > Members > API tab)
3. Click Create API key
4. Copy the key (format: sk-gamma-xxxxxxxx)

Note: Requires Gamma Pro, Ultra, Team, or Business account.

3. Configure Your AI Assistant

Choose your AI assistant below for setup instructions.

Claude Desktop (macOS / Windows / Linux)

Config File

Add to the mcpServers object:

{
  "mcpServers": {
    "gamma": {
      "command": "node",
      "args": ["/absolute/path/to/gamma-mcp-server/dist/index.js"],
      "env": {
        "GAMMA_API_KEY": "sk-gamma-your-api-key-here"
      }
    }
  }
}

Restart Claude Desktop

Restart Claude Desktop to load the new MCP server. You should see "gamma" in your MCP servers list.

Claude Code

Claude Code uses the same MCP configuration as Claude Desktop. If you've already configured Claude Desktop, you're all set.

For project-level configuration, create a .claude/settings.json file in your project directory with the same mcpServers structure shown above. This allows different projects to use different MCP server configurations.

OpenCode

Config File

JSON Structure

{
  "mcp": {
    "gamma": {
      "type": "local",
      "command": ["node", "/absolute/path/to/gamma-mcp-server/dist/index.js"],
      "enabled": true,
      "environment": {
        "GAMMA_API_KEY": "sk-gamma-your-api-key-here"
      }
    }
  }
}

Note: OpenCode uses a different config format from Claude Desktop — mcp (not mcpServers), type field required ("local" or "remote"), command is an array, and env vars go under environment.

Restart OpenCode after editing the config to load the server.

GitHub Copilot CLI

Config File

• ~/.copilot/mcp-config.json

JSON Structure

{
  "mcpServers": {
    "gamma": {
      "type": "local",
      "command": "node",
      "args": ["/absolute/path/to/gamma-mcp-server/dist/index.js"],
      "env": {
        "GAMMA_API_KEY": "sk-gamma-your-api-key-here"
      },
      "tools": ["*"]
    }
  }
}

Note: Requires the GitHub Copilot CLI (gh copilot) — not the same as OpenAI Codex.

OpenAI Codex

Config File

• ~/.codex/config.toml (TOML format, not JSON)

TOML Structure

[mcp_servers.gamma]
command = "node"
args = ["/absolute/path/to/gamma-mcp-server/dist/index.js"]
enabled = true

[mcp_servers.gamma.env]
GAMMA_API_KEY = "sk-gamma-your-api-key-here"

Note: Codex uses TOML format, not JSON. The env section is a separate table under [mcp_servers.gamma.env].

Google Gemini CLI

Config File

JSON Structure

{
  "mcpServers": {
    "gamma": {
      "command": "node",
      "args": ["/absolute/path/to/gamma-mcp-server/dist/index.js"],
      "cwd": "/absolute/path/to/gamma-mcp-server",
      "env": {
        "GAMMA_API_KEY": "sk-gamma-your-api-key-here"
      },
      "timeout": 30000
    }
  }
}

Restart Gemini CLI after editing the config to load the server.

Available Tools

gamma_generate

Create new content using Gamma's AI.

Formats & Sizes:

• presentation: fluid, 16x9, 4x3
• document: fluid, pageless, letter, a4
• social: 1x1, 4x5, 9x16
• webpage: fluid

Example prompts in your AI assistant:

• "Create a 5-slide presentation about sustainable energy"
• "Generate a document explaining our Q1 results"
• "Make a social media post announcing our new product"

gamma_get_status

Check if a generation has completed. Set waitForCompletion: true to automatically poll until done.

gamma_from_template

Remix an existing Gamma with new content or variable substitutions.

{
  "templateId": "gamma_xyz789",
  "prompt": "Update for Q1 2025",
  "variables": { "company_name": "Acme Corp" }
}

Multi-Machine Setup

This repository is designed for easy deployment across multiple machines:

# On each machine:
git clone https://github.com/Arkava-AI/gamma-mcp-server.git
cd gamma-mcp-server
npm install && npm run build

# Then configure your AI assistant with the local path

To update on any machine:

git pull
npm install
npm run build
# Restart your AI assistant

Environment Variables

Development

# Run in development mode with auto-reload
npm run dev

# Build for production
npm run build

# Run linting
npm run lint

# Format code
npm run format

# Type check
npm run typecheck

# Test with MCP Inspector
npm run inspect

Project Structure

gamma-mcp-server/
├── src/
│   ├── index.ts          # Main entry point
│   ├── constants.ts      # Configuration constants
│   ├── types.ts          # TypeScript interfaces
│   ├── schemas/          # Zod validation schemas
│   ├── services/         # API client and formatters
│   └── tools/            # MCP tool implementations
├── dist/                 # Compiled JavaScript (generated)
├── package.json
├── tsconfig.json
└── eslint.config.js

API Credits

Gamma uses a credit-based system for API usage. Credits are consumed per generation. Monitor your usage in the Gamma dashboard and enable auto-recharge if needed.

Troubleshooting

Requirements

• Node.js 18+
• Gamma Pro/Ultra/Team/Business account (for API access)
• MCP-compatible AI assistant (Claude Code, Claude Desktop, OpenCode, GitHub Copilot CLI, Gemini CLI, etc.)

Maintainer

Arkava Ltd — engage@arkava.ai

License

MIT License - see LICENSE for details.

Links

• Gamma Developer Docs
• MCP Specification
• Gamma.app