Meta Ads MCP Server
A comprehensive MCP server for managing and analyzing Meta Ads (Facebook/Instagram) with over 80 natural-language tools for AI agents like Claude Desktop.
README
title: Meta Ads MCP Server emoji: 📣 colorFrom: blue colorTo: indigo sdk: docker pinned: false license: mit app_port: 7860 tags:
- mcp
- meta-ads
- facebook-ads
- marketing-api
- llm-tools
Meta Ads MCP Server
A comprehensive Model Context Protocol (MCP) server for managing and analyzing Meta Ads (Facebook/Instagram). This server exposes over 80 natural-language tools to AI agents like Claude Desktop.
🚀 Quick Start
To connect your Claude Desktop to this deployed server:
- Open your Claude Desktop configuration file:
- Mac:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\\Claude\\claude_desktop_config.json
- Mac:
- Add the following configuration (replace
<username>with your HF username):
{
"mcpServers": {
"meta-ads": {
"type": "sse",
"url": "https://<username>-meta-ads-mcp.hf.space/sse"
}
}
}
- Restart Claude Desktop. You should now see the
meta-ads-mcptools available.
⚙️ Setup Guide (Hugging Face Spaces)
To deploy your own instance of this server:
- Create a Meta App:
- Go to Meta for Developers
- Create an App (Type: Business or Marketing)
- Note the App ID and App Secret.
- Generate a Long-Lived System User Access Token:
- Go to Business Settings -> System Users
- Assign the user to your Ad Account and hit "Generate Token".
- Add the Hugging Face Repository Secrets (Settings -> Settings -> Repository Secrets):
META_ACCESS_TOKEN: The token you just generated.META_APP_ID: Your App ID.META_APP_SECRET: Your App Secret.DEFAULT_AD_ACCOUNT: Formatact_XXXXXXXXXXXXX.META_BUSINESS_ID: (Optional) For catalog/dataset creation.
🔐 Required Permissions
Your Meta Access Token MUST have the following permissions granted:
ads_read(For reading campaigns/insights)ads_management(For creating/editing ads)business_management(For account lookups)pages_read_engagement(For page validations)pages_manage_ads(For creatives)instagram_basic&instagram_content_publish(For IG creatives)catalog_management(For product feeds)read_insights(For analytics tools)
🛠️ Tool Reference
The server exposes an exhaustive list of tools divided into modules:
- Campaigns:
create_campaign,update_campaign,delete_campaign,duplicate_campaign... - Ad Sets & Targeting:
create_adset,build_targeting_spec,search_targeting_interests... - Ads:
create_ad,get_ad_preview,duplicate_ad... - Creatives:
create_image_creative,create_video_creative,upload_image_from_file... - Audiences:
create_lookalike_audience,create_website_audience,share_audience... - Insights:
get_account_insights,get_campaign_insights,create_async_report... - Pixels:
send_conversion_event,get_pixel_stats... - Catalogs:
create_product_feed,list_products...
Note: For a full list of tools, request Claude to "List all available Meta Ads tools".
💡 Example Prompts
Try these prompts in Claude Desktop once connected:
- "Create a Traffic campaign targeting 25-45 year olds in India interested in technology, with a ₹5,000/day budget."
- "Show me the performance of all active campaigns in the last 30 days, broken down by age and gender."
- "Pause all ad sets spending more than $100/day with a CTR below 0.5%."
- "Create a 3% lookalike audience of my website visitors from the last 60 days, targeting the US and UK."
- "Upload this product image and create a carousel ad with 5 cards linking to different product pages."
- "Run an async report of all campaign performance for Q1 2025, broken down by placement and device."
- "Send a server-side purchase conversion event for order ID 9821 worth $149.99."
- "Build a targeting spec for men 30-55 in California interested in golf, luxury cars, and travel."
- "Show me which of my creatives have the highest video completion rate this month."
- "Duplicate my best-performing campaign from last quarter with a 20% higher budget."
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。