pipeboard-co/meta-ads-mcp

What it solves

This project provides a way for AI assistants (like Claude, ChatGPT, and Cursor) to manage Meta Ads campaigns end-to-end through natural language. It eliminates the need for marketers to manually navigate the Meta Ads Manager interface by allowing them to launch campaigns, upload creatives, and analyze performance via a conversational interface.

How it works

It is a Model Context Protocol (MCP) server that connects AI clients to Meta's public APIs. It can be deployed as a hosted remote MCP—meaning no self-hosting or developer tokens are required—or installed locally for technical users. The server exposes a set of tools (such as create_campaign and get_campaigns) that the AI agent can call to execute actions in the Meta Ads account.

Who it’s for

• Digital Marketers: Who want to use AI to automate and analyze their advertising efforts on Facebook and Instagram.
• AI Power Users: Who use MCP-compatible clients like Claude Desktop or Cursor to integrate external tools into their LLM workflows.
• Developers: Who want to integrate Meta Ads management capabilities into their own AI agents or applications.

Highlights

• End-to-End Management: Full control over campaigns, ad sets, ads, and creatives.
• Remote MCP Hosting: Quick setup via a hosted URL without needing to manage infrastructure.
• Cross-Platform Ecosystem: Part of the Pipeboard MCP family, sharing authentication and safety models across Meta, Google, TikTok, Snap, and Reddit Ads.
• Safety Guardrails: Explicit confirmation is required for every write action, and new campaigns start paused by default.
• Broad Toolset: Includes 42 specific tools for targeting, budget optimization, and performance insights.