Skip to main content
Connect your AI assistant to Datafold to run data diffs, query data sources, and manage monitors.

Connect

Server URL: https://app.datafold.com/mcp
  • Claude Code: claude mcp add --transport http datafold https://app.datafold.com/mcp
  • Claude (claude.ai or Claude Desktop): add a custom connector with the server URL.
Then follow your client’s sign-in prompts. For other clients, see the setup guide.
Replace app.datafold.com with the host you log in to: eu.datafold.com in the EU, or your Classic Dedicated Cloud domain.
To revoke a client’s access, go to Settings > Account > Authorized OAuth apps.

API key

For MCP clients that do not support OAuth:
1

Get your API key

Go to Settings > Account in the Datafold app and click Create API Key.
2

Connect your AI assistant

For Claude Code, run:
For Cursor, create .cursor/mcp.json in your project:
For other AI assistants, see the full setup guide.

Tool Visibility and Permissions

The tools available to your AI agent depend on the permissions of the user who signed in, or of the API key’s owner. Tools that require permissions the user doesn’t have are automatically hidden from the agent. To scope an agent to a specific set of tools, create a custom group with only the permissions you need, assign it to a service account, and use that service account’s API key. See MCP Tool Permissions for the exact permissions each MCP tool requires.

Supported Clients

Datafold’s MCP server works with any client that supports the Model Context Protocol, including Claude Desktop, Claude Code, Cursor, VS Code with Cline, Windsurf, Continue.dev, Zed, OpenCode, Gemini CLI, and Kiro.

Full MCP Setup Guide

Detailed setup instructions for all supported clients, troubleshooting, and best practices.