ArmaVita Meta Ads MCP
A local Model Context Protocol server that enables interaction with the Meta Marketing API to manage ad accounts, campaigns, and creatives. It provides tools for targeting research, insight reporting, and campaign management through local MCP clients like Claude Code and Cursor.
README
armavita-meta-ads-mcp
<p align="center"> <img src="docs/assets/armavita-meta-ads-mcp-hero-1080.jpg" alt="ArmaVita Meta Ads MCP hero image" width="100%" /> </p>
<p align="center"><strong>Brought to you by <a href="https://armavita.com">ArmaVita.com</a></strong></p> <p align="center">Need a custom implementation? <a href="https://armavita.com">Contact us</a>.</p>
armavita-meta-ads-mcp is a local Model Context Protocol server for Meta Ads.
It is built for local MCP clients (Claude Code, Cursor, Codex) and supports:
- Meta access token auth (
META_ACCESS_TOKEN) - Local OAuth flow (
META_APP_ID+META_APP_SECRET) - stdio MCP transport only
- Python
3.11+ mcp[cli]==1.26.0- Meta Marketing API
v25.0by default (META_GRAPH_API_VERSIONoverride supported)
Current contract version: 1.1.0.
Install
From PyPI (once published):
pip install armavita-meta-ads-mcp
From source (recommended during development):
uv sync
Run
armavita-meta-ads-mcp
Module entrypoint:
python -m armavita_meta_ads_mcp
Login flow:
armavita-meta-ads-mcp --login
Quick MCP Client Config
Minimal MCP server registration (JSON format used by many clients):
{
"mcpServers": {
"meta-ads-armavita": {
"command": "armavita-meta-ads-mcp",
"env": {
"META_ACCESS_TOKEN": "EA...",
"META_GRAPH_API_VERSION": "v25.0"
}
}
}
}
OAuth mode (no direct token in config):
{
"mcpServers": {
"meta-ads-armavita": {
"command": "armavita-meta-ads-mcp",
"env": {
"META_APP_ID": "YOUR_APP_ID",
"META_APP_SECRET": "YOUR_APP_SECRET"
}
}
}
}
Then run once to complete login:
armavita-meta-ads-mcp --login
Tool Coverage
- Accounts:
list_ad_accounts,read_ad_account - Campaigns:
list_campaigns,read_campaign,create_campaign,update_campaign - Budget schedules:
create_campaign_budget_schedule - Ad sets:
list_ad_sets,read_ad_set,create_ad_set,update_ad_set - Ads/creatives/media:
list_ads,read_ad,list_ad_previews,create_ad,update_ad,list_ad_creatives,read_ad_creative,create_ad_creative,update_ad_creative,upload_ad_image_asset,read_ad_image,export_ad_image_file,search_pages,list_account_pages - Insights/reporting:
list_insights,create_report - Targeting:
search_interests,suggest_interests,estimate_audience_size,search_behaviors,search_demographics,search_geo_locations - Duplication:
clone_campaign,clone_ad_set,clone_ad,clone_ad_creative - Ads Library:
search_ads_archive - Research helpers:
search_web_content,read_web_content
Pagination
Cursor-based pagination is supported on core list/read streams:
list_ad_accounts,list_campaigns,list_ad_sets,list_ads,list_insightslist_ad_creatives,search_interests,suggest_interests,search_behaviors,search_demographics,search_geo_locations,search_ads_archive- Use
page_cursorwith thepaging.cursors.aftervalue from the previous response. - Responses preserve Meta's native
pagingobject.
Insights Query Notes
list_insightsandcreate_reportsupport either:date_rangeas{ "since": "YYYY-MM-DD", "until": "YYYY-MM-DD" }, ordate_rangeas a preset (for examplelast_30d,maximum).
create_report.comparison_perioduses the same format and validation asdate_range.previous_30dis normalized tolast_30d.- For action metrics, use
action_breakdowns(and optionalsummary_action_breakdowns) instead of mixing action keys intobreakdowns.
Security
- Access tokens are redacted from URL fields returned by the server (including nested
paging.nextURLs).
Docs
Scope
- This repository is an OSS local MCP server.
- Transport mode is local
stdioonly. - Tool aliases are intentionally not exposed.
License
GNU Affero General Public License v3.0 (AGPLv3). See LICENSE.
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。