# Windsor.ai MCP Server

> Windsor.ai MCP Server connects AI agents to 325+ marketing, advertising,
> analytics, e-commerce, CRM, and database sources via the Model Context
> Protocol. Query live data from Meta Ads (Facebook & Instagram), Google Ads,
> TikTok Ads, LinkedIn Ads, Microsoft Ads (Bing), Snapchat Ads, Pinterest Ads,
> Google Analytics 4, HubSpot, Salesforce, Klaviyo, Shopify, Stripe, BigQuery,
> Snowflake and more — directly from Claude, ChatGPT, Cursor, Copilot, or any
> MCP-compatible client.

## Quick Start

- MCP endpoint (Streamable HTTP): https://mcp.windsor.ai/
- MCP endpoint (SSE): https://mcp.windsor.ai/sse/
- Authentication: OAuth 2.0 (recommended) or Windsor.ai API key (Bearer token)

## Documentation

- [Full reference for LLMs](https://mcp.windsor.ai/llms-full.txt)
- [HTML documentation](https://mcp.windsor.ai/docs)
- [Available data sources (JSON)](https://mcp.windsor.ai/datasources)
- [OAuth metadata](https://mcp.windsor.ai/.well-known/oauth-authorization-server)
- [Windsor.ai homepage](https://windsor.ai/)

## Available Tools

- `get_current_user`: Return the authenticated user's profile (username, email).
- `get_connectors`: List Windsor.ai connectors and their connected accounts.
  Pass `include_not_yet_connected=True` to also see connectors the user can set up.
- `get_connector_authorization_url`: Get a browser link to connect or authorize
  a connector (works for OAuth and manual credential connectors).
- `get_options`: List fields, date filters and connector-specific options for
  a given connector and account set.
- `get_fields`: Get descriptions, types and tables for specific fields.
- `get_data`: Run a query against a connector with date ranges, filters,
  options and account selection.

## Compatible Clients

Claude (Desktop, Web, Code), ChatGPT, Cursor, Windsurf, Cline, GitHub Copilot,
Google Gemini, Microsoft Copilot, Manus, n8n, mcp-proxy, and any application
built on the MCP open standard.