Adops MCP
概览
作者提供的说明(英文原文)
CSV-first analytics for Google Ads + Meta Ads: reporting, anomaly detection, budget optimization.
PopNx MCP 的三大支柱
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
- 用户评价i暂无评价评价功能将在后续阶段与账户功能一起开放
- 我们的基础检查i1 项中通过 1 项依据公开信息检查;尚未扫描代码
- 使用情况i非常新↓ 91/周 · ★ –
它能帮你做什么
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
- 连接方式: 在你的电脑上运行
- 传输方式: stdio
- 分类: 营销与社交媒体
安装
选择你使用的 AI 应用,然后复制设置。
安装之前
- 基础检查:1 项中通过 1 项
- 在你的电脑上运行,拥有你的用户账户的权限(文件、网络、应用)。
- 注册中心没有列出需要填写的密钥。
你需要
- npm已安装 Node.js(用于 npx 命令)
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"adops": {
"command": "npx",
"args": [
"-y",
"adops-mcp-server"
]
}
}
}在终端中运行这条命令
claude mcp add adops -- npx -y adops-mcp-server或者自己复制设置
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"adops": {
"command": "npx",
"args": [
"-y",
"adops-mcp-server"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
或者自己复制设置
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"adops": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"adops-mcp-server"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它能做什么
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
基础检查i
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日
- ✓ 通过
作者与仓库一致
注册名称已通过 GitHub 账户 enzoemir1 验证,该账户拥有此仓库
- … 尚未检查
仍在维护
我们还没有从 GitHub 读取这个仓库
- … 尚未检查
有许可证
我们还没有从 GitHub 读取这个仓库
- – 不适用
端点有响应
它在你的电脑上运行,没有互联网端点
- – 不适用
使用 OAuth 登录
在你的电脑上运行,改用本地设置
- … 尚未检查
工具描述中没有隐藏指令
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与“营销与社交媒体”中的其他服务器比较i
- 星标数多于此分类中其他 3611 个服务器的 0%
- 发布到注册中心的时间比此分类中其他 3611 个服务器的 6% 更新
- 基础检查结果等于或优于此分类中其他 3611 个服务器的 100%
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
常见问题
Adops MCP 安全吗?
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
它需要哪些密钥或代码?
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在我的电脑上运行还是在云端?
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
它还在维护吗?
GitHub 仓库最后更新于–,最新版本(1.3.0)于7个月前进入注册中心。
哪些应用可以使用它?
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。
有什么可以替代它?
同一分类中有:MCP YouTube Transcript, Google Analytics MCP, BibiGPT YouTube Video Summarizer
替代选择
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Google Analytics MCP
GA4 MCP server for AI agents: schema discovery, server-side aggregation, analysis-ready defaults.
BibiGPT YouTube Video Summarizer
AI video summarizer: summarize and transcribe YouTube, Bilibili, TikTok and podcast videos.
作者的 README
打开 README
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
AdOps MCP
CSV-first analytics & optimization for Google Ads + Meta Ads — inside Claude, Cursor, and any MCP client.
Export a CSV from Google Ads or Meta Ads Manager → import into AdOps → ask your AI assistant anything about your campaigns.
No OAuth. No developer tokens. No API approval queues. Your data stays on your machine.
Why CSV-first?
Google Ads API and Meta Marketing API are powerful but gated: developer token approval takes days, Meta app review takes weeks, OAuth flows are fragile. For most advertisers those barriers never clear.
AdOps takes a different path. Every ad platform's dashboard already has a Download / Export button that produces a CSV of exactly what you care about: campaign performance, day-segmented metrics, spend, conversions, ROAS. AdOps reads that CSV and runs the same analysis you'd get from a $200/mo optimization suite — anomaly detection, budget allocation, A/B significance, forecasting, industry benchmarks — directly inside your AI assistant.
You stay in control of your data. You don't hand credentials to a third party. You don't wait 3 weeks for a review.
Features
- 17 MCP tools covering the full ad analytics lifecycle
- 4 MCP resources for at-a-glance dashboards
- CSV import from Google Ads Editor, Google Ads web UI, and Meta Ads Manager — with automatic format detection (US vs European number formatting, CRLF vs LF, BOM handling)
- Unified cross-platform schema — Google and Meta campaigns normalized into a single
UnifiedCampaign/UnifiedMetricsmodel - AI-powered recommendations — budget reallocation, scaling high-ROAS campaigns, pausing zero-conversion ones
- Bulk-edit CSV export — turn recommendations into Google Ads Editor / Meta Ads Manager CSVs you can paste straight into the dashboard for one-click bulk updates
- Statistical anomaly detection — CPC spikes, CTR drops, spend surges, sensitivity-configurable
- A/B test analysis with confidence scoring
- 9-vertical industry benchmarks built-in (eCommerce, SaaS, B2B, etc.)
- Spend & conversion forecasting (7/14/30 days)
- Demo mode — seed a realistic portfolio to try the tools before importing your own data
- 91 automated tests — TypeScript strict mode, Zod validation throughout
Quick Start
npm i adops-mcp-server
Add to your MCP client (Claude Desktop claude_desktop_config.json, Cursor, VS Code, Windsurf):
{
"mcpServers": { "adops": { "command": "node", "args": ["path/to/node_modules/adops-mcp-server/dist/index.js"] } } }
Restart your client. Then, inside Claude:
> Seed demo data so I can explore AdOps
(calls ad_demo_seed)
Or import your real data:
> Import ~/Downloads/google-ads-report.csv as a Google Ads CSV
(calls ad_csv_import with platform=google, csv_path=...)
Exporting CSV from Your Ad Dashboard
Google Ads
- Open ads.google.com → Campaigns view
- (Optional) Click Segment → Day in the toolbar for daily metrics
- Click Download (⬇) icon → choose .csv format
- Save to disk, give the absolute path to
ad_csv_import
Meta Ads Manager
- Open business.facebook.com → Ads Manager
- Switch to Campaigns tab (not Ad Sets or Ads — campaign-level rollups)
- (Optional) Breakdown → Time → Day for daily metrics
- Click Reports → Export → Campaign performance (.csv)
- Save to disk, give the absolute path to
ad_csv_import
Both exports work out-of-the-box with or without day segmentation. Without a Day column, you get one aggregated metric row per campaign. With Day segmentation, you get daily time-series data — which powers forecasting and anomaly detection properly.
Tools
| Tool | Description |
|---|---|
ad_demo_seed | Seed a realistic cross-platform demo portfolio (8 campaigns, 30 days of metrics) to try AdOps before importing real data |
ad_csv_import | Import a Google Ads or Meta Ads CSV export — every other tool operates on this data |
ads_export_recommendations | Export budget_analyze recommendations as Google Ads Editor CSV, Meta Ads Manager CSV, JSON, or Markdown — one bulk-paste and your ad dashboard reflects the AI's plan |
platform_connect | Register a data source connection manually (alternative to CSV import, for scripted workflows) |
campaign_list | List and filter campaigns across all imported platforms |
campaign_create | Add a campaign manually (useful for what-if planning) |
campaign_update | Update campaign settings in local storage (budget, status, bidding, schedule) |
campaign_pause_resume | Batch pause or resume up to 50 campaigns |
ads_report | Generate unified cross-platform performance report |
budget_analyze | Analyze budget allocation, produce scaling/pausing recommendations |
budget_reallocate | Transfer budget between campaigns across platforms |
audience_insights | Demographic / geographic / device breakdowns |
creative_specs | Platform-specific image, video, and text requirements reference |
anomaly_detect | Detect performance anomalies with configurable sensitivity |
ab_test_analyze | Compare two campaigns with statistical significance testing |
competitor_benchmark | Compare your metrics against industry averages (9 verticals) |
forecast_spend | Forecast spend, conversions, ROAS for the next 7/14/30 days |
Resources
| Resource | Description |
|---|---|
ads://overview | Cross-platform dashboard summary |
ads://campaigns | All active campaigns with key metrics |
ads://budget | Budget allocation across platforms |
ads://alerts | Recent performance anomalies and warnings |
Unified Metrics Schema
AdOps normalizes metrics across platforms:
| Metric | Formula | Notes |
|---|---|---|
| CTR | clicks / impressions × 100 | Click-through rate (%) |
| CPC | spend / clicks | Cost per click |
| CPM | spend / impressions × 1000 | Cost per 1000 impressions |
| ROAS | conversion_value / spend | Return on ad spend |
| CPA | spend / conversions | Cost per acquisition |
| Conversion Rate | conversions / clicks × 100 | % |
CSV column mapping:
| AdOps Field | Google Ads CSV | Meta Ads CSV |
|---|---|---|
spend | Cost (USD equivalent auto-converted if micros) | Amount spent (USD/EUR/...) |
impressions | Impressions | Impressions |
clicks | Clicks | Link clicks |
conversions | Conversions | Results (Purchases/Leads/etc.) |
conversion_value | Conv. value | Purchases conversion value |
reach | — | Reach (Meta-only) |
quality_score | Quality Score (if present) | — (Google-only) |
European locale numbers (125,50 = 125.50) and semicolon-delimited CSVs are auto-detected.
Configuration
AdOps is stateless from the network's point of view — no API keys, no OAuth. The only configuration is where it persists imported data:
| Variable | Description | Default |
|---|---|---|
ADOPS_DATA_DIR | Directory for JSON persistence | ./data |
All other behavior is controlled via tool arguments.
Pricing
| Tier | Price | Features |
|---|---|---|
| Free | $0 | 1 imported CSV per day, demo mode, all 17 tools read-only |
| Pro | €24 lifetime | Unlimited imports, full CRUD, all analytics tools |
| Agency | $59/mo | Multi-account, white-label reports, priority support |
Available via GitHub (self-hosted, free) or the MCPize marketplace (managed).
Development
git clone https://github.com/enzoemir1/adops-mcp.git
cd adops-mcp npm ci npm run build npm test # 91 tests across 10 suites npm run inspect # Open MCP Inspector
Tests
91 tests covering:
- CSV parser (11 tests): RFC 4180 quoting, BOM, European/US number formats, delimiter detection
- Google Ads CSV import (7 tests): campaigns report, day segmentation, objective mapping, European locale
- Meta Ads CSV import (5 tests): Meta-specific columns, currency detection from header, objective mapping, reach/frequency capture
- Export recommendations (8 tests): Google Ads Editor CSV format, Meta Ads Manager CSV format, JSON wrapper, Markdown rendering, min_delta_pct filter, limit, by_type summary
- Storage: Connection CRUD, campaign search, metrics aggregation, batch inserts
- Analytics: Metric calculations, performance reports, forecasting, benchmarks
- Optimizer: Budget analysis, reallocation, scaling/pausing recommendations
- Anomaly: CPC spike detection, conversion drops, sensitivity, severity sorting
- E2E Workflow: 14 real user scenarios
Roadmap
- v1.3 — Add Google Ads Editor
.csv.gzand Excel.xlsxsupport - v1.4 — Live Google Ads / Meta Ads API integration as an optional enterprise feature (BYOK — bring your own developer token + app review)
- v2.0 — Multi-user server mode with JWT auth for team deployments
Pro License
AdOps ships in Free mode — ad_demo_seed, ad_csv_import, platform_connect, campaign_list, campaign_create, campaign_update, campaign_pause_resume, ads_report, budget_reallocate, audience_insights, and creative_specs are open. The following tools require a Pro license:
ads_export_recommendations— bulk-edit CSV export for Google Ads Editor / Meta Ads Managerbudget_analyze— AI budget allocation analysis with action recommendationsanomaly_detect— statistical anomaly detectionab_test_analyze— A/B test z-test with significance + upliftforecast_spend— 7/14/30-day spend & ROAS forecastcompetitor_benchmark— industry benchmark comparison
Buy a Pro License (€24, lifetime, 3 machines): https://automatiabcn.lemonsqueezy.com/buy/1525c929-832c-4472-a88a-58edbfa4e87b
Or get the [Indie MCP Stack Bundle](https://automatiabcn.lemonsqueezy.com/buy/55e932fd-8319-47f0-8e95-0b86a29f2617) (€69, all 4 servers).
export LEMONSQUEEZY_LICENSE_KEY=YOUR-KEY-HERE
Or in your MCP client config:
{
"mcpServers": { "adops-mcp": { "command": "npx", "args": ["-y", "adops-mcp-server"], "env": { "LEMONSQUEEZY_LICENSE_KEY": "YOUR-KEY-HERE" } } } }
Validation is cached locally for 24 h — fully offline-capable after first run.
License
MIT License. See LICENSE for details.
Built by Automatia BCN.