# Windsor.ai MCP

Connects an AI agent to marketing, sales, and business data across 325+ platforms. Use it for reports, campaign analysis, and supported write actions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 7
- Author: windsor-ai
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/windsor-ai/windsor_mcp
- Updated: 2026-08-19
- HTML: https://affpapa.org/en/nevibecode/mcp/windsorai-mcp

## What it does

Windsor.ai MCP lets an AI client query live business data through MCP. It can inspect advertising performance, join data from several connected services, and prepare sales or revenue reports without hand-written SQL. Available sources include Meta Ads, Google Ads, TikTok Ads, GA4, HubSpot, Salesforce, Shopify, Stripe, BigQuery, and Snowflake.

The server uses `https://mcp.windsor.ai/sse` and OAuth 2.0. On the first connection, the client sends the user to Windsor.ai to sign in, so there is no API key to manage. Some connectors expose write actions, such as changing ad campaigns, publishing posts, or updating Salesforce records. Other connectors are read-only.

Queries run inside the client chat and return structured results. Since write actions change external state, the client should confirm the user’s intent before running them.

## Requirements
- Remote server: `https://mcp.windsor.ai/sse`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport sse windsor-mcp https://mcp.windsor.ai/sse
```

### Codex

**Codex**
- Codex connects MCP servers over stdio and streamable HTTP only, and this server speaks SSE

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "windsorai-mcp": {
            "url": "https://mcp.windsor.ai/sse"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "windsorai-mcp": {
            "type": "sse",
            "url": "https://mcp.windsor.ai/sse"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t sse windsorai-mcp https://mcp.windsor.ai/sse
```

## When to use it
- build a report on advertising spend by channel
- compare campaigns by ROAS and identify weak performers
- combine GA4 sessions with Google Ads spend
- pull Shopify orders and Stripe payments for a period
- pause an ad campaign through a supported connector

## Example prompts
- Compare Meta Ads campaigns by ROAS and spend for last month.
- Combine GA4 and Google Ads data into a weekly report.
- Find low-performing campaigns and suggest which ones to pause.

## Similar
- [windsor-ai-business-data](https://affpapa.org/en/nevibecode/skills/windsor-ai-business-data.md) — skill · 845 installs
- [yandex-metrica-mcp](https://affpapa.org/en/nevibecode/mcp/yandex-metrica-mcp.md) — MCP server · ★ 15
- [Yandex Metrika MCP](https://affpapa.org/en/nevibecode/mcp/yandex-metrika-mcp.md) — MCP server · ★ 19
- [performance-analyzer-sms](https://affpapa.org/en/nevibecode/skills/performance-analyzer-sms.md) — skill · 1700 installs
- [google-analytics](https://affpapa.org/en/nevibecode/skills/google-analytics.md) — skill · 520 installs
- [ad-attribution-gap](https://affpapa.org/en/nevibecode/skills/ad-attribution-gap.md) — skill · 20500 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md