How to integrate data into Claude with Windsor MCP
Windsor MCP for Claude allows you to unlock AI-powered insights and visual summaries from your business data across 325+ platforms in seconds.
Follow this guide to connect the Windsor MCP server with your Claude account using the official connector.
Prerequisites
To use Windsor MCP with Claude, make sure you have the following:
- An active Windsor.ai account (free or paid plan) with at least one connected data source
- Claude account
🚀 Get started with Windsor MCP today with a 30-day free trial: https://onboard.windsor.ai/ and experience the power of AI-driven analytics.
Steps to configure Windsor MCP for Claude
Method 1. Using Claude Desktop
1. In Claude, go to the Connectors page, browse ‘Windsor.ai’ and add it using the ‘+‘ button.

2. You’ll be prompted to authorize the connection by signing in with your Windsor.ai account.
3. When done, go back to Claude. Windsor MCP will appear in the list of connectors.

For quicker connector access and use, you can choose ‘Always allow‘ permissions.

4. Now, create a new Claude chat and feel free to ask any questions related to your data.

Method 2. Developer Mode Setup
If you want more control over your data, you can configure Windsor MCP using the dev tools provided.
1. Install the Windsor MCP Proxy:
Run this command in the terminal to find the exact path:
- Open Claude Desktop → Settings → Developer → Edit Config
This opens a JSON config file. Add the following:
{
"mcpServers": {
"windsor": {
"command": "mcp-proxy", // path to your mcp-proxy
"args": [
"https://mcp.windsor.ai/",
"--transport=streamablehttp"
],
"env": {
"API_ACCESS_TOKEN": "<your-windsor-api-key>"
}
}
}
}
4. Restart Claude Desktop and start querying:
Quit and reopen Claude. You should see “windsor” listed in your MCP options. Feel free to ask the first question related to your data.
Windsor MCP connector functions
The Windsor MCP connector provides four core functions that allow you to discover connectors, explore available fields, and retrieve structured marketing data for AI analysis and reporting:
- get_connectors
- get_options
- get_fields
- get_data
Function overview
| Function | Purpose | What it returns | When to use it |
|---|---|---|---|
| get_connectors | Retrieve available connectors and connected accounts | List of supported data sources (e.g., Facebook, Google Ads, TikTok) and linked accounts | When initializing setup or verifying connected accounts |
| get_options | Retrieve available fields, date filters, and connector-specific options | Available metrics, dimensions, attribution settings, and date configurations | Before building a query to understand what data is available |
| get_fields | Get detailed metadata about specific fields | Field descriptions, types, and technical details | When validating metrics/dimensions or building advanced queries |
| get_data | Retrieve structured marketing data | Performance data filtered by fields, date ranges, filters, and attribution settings | When pulling data for dashboards, reports, or AI analysis |
Helpful links
- About Windsor MCP – Explore general info about Windsor MCP: supported data sources and LLMs, benefits, use cases, etc.
- Windsor MCP use cases & examples (30+ prompts inside) – If you don’t know what to ask or are looking for advanced Windsor MCP use cases, read this guide for a batch of practical examples and useful prompts to make your Claude data analysis even more effective and deeper.
- Connect Facebook Ads to Claude – Visit this sample integration page to see how you can connect a specific data source (Meta Ads) to Claude and what benefits it provides.
Tired of manual data reporting and analysis? Get started with Windsor MCP today to unlock instant AI insights

Windsor vs Coupler.io