This repository provides a unified MCP (Model Context Protocol) server designed to streamline campaign management, analytics, and optimization across Google Ads, Meta Ads, and Google Analytics 4. It targets marketers, agencies, and developers seeking to leverage AI assistants like ChatGPT and Claude for comprehensive ad operations. The primary benefit is consolidating disparate platform management into a single, remotely accessible server, significantly reducing setup complexity and enabling data-driven decision-making.

How It Works

The core innovation is a single, remote MCP server that acts as an intermediary, abstracting the complexities of individual platform APIs (Google Ads, Meta Ads, GA4). AI assistants connect to this server, which then routes requests and data between the AI and the respective advertising and analytics platforms. This approach offers full read and write capabilities, unlike some official tools, and integrates GA4 data for holistic campaign analysis. Authentication is handled via industry-standard OAuth 2.1 with PKCE, ensuring secure and scoped access without storing sensitive credentials on the server.

Quick Start & Requirements

Setup involves configuring your chosen AI assistant (e.g., Claude Desktop, ChatGPT, n8n) to connect to a provided MCP endpoint URL (YOUR_MCP_ENDPOINT_URL). No direct server installation is required for the user; the focus is on client integration. Prerequisites include access to compatible AI assistants and connecting your Google and Meta ad accounts via OAuth prompts. Detailed setup guides are available for various assistants, including Claude Desktop, ChatGPT, and n8n.

Highlighted Details

• Offers over 250 tools covering Google Ads (150+), Meta Ads (80+), and GA4 (20+).
• Provides full read and write access across all managed platforms.
• Enables cross-platform performance comparison, unified ROAS analysis, and budget allocation recommendations.
• Supports integration with a wide array of AI assistants, including ChatGPT, Claude, Cursor, n8n, and Windsurf.
• Features a remote server model for rapid, 2-minute setup.

Maintenance & Community

The project welcomes contributions via issue templates for bug reports and feature requests. While specific community channels like Discord or Slack are not detailed, the repository encourages user engagement through its issue tracker and star history.

Licensing & Compatibility

This project is licensed under the permissive MIT License, allowing for broad compatibility with commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

The MCP server itself does not store user ad data or credentials; authentication tokens remain local to the user's machine. Write operations require explicit user confirmation, acting as a safety measure rather than a functional limitation. The README does not detail specific known bugs or platform version limitations.