RudderStack
Customer data pipeline inspection, debugging, and configuration changes from tools like Claude Desktop and Cursor
ON THIS PAGE
How to Connect to RudderStack AI
Connect your AI assistant or development tool to RudderStack AI for secure and efficient pipeline debugging.
This guide will help you connect your AI assistant or development tools to RudderStack’s MCP (Model Context Protocol) server. The RudderStack MCP server provides a standardized interface that allows any compatible AI model or agent to access your RudderStack data securely.
This is an early access feature. If you encounter any issues or have suggestions for improvement, reach out to RudderStack Support.
Prerequisites
- Active RudderStack account with workspace access
- Compatible MCP client (Claude, VS Code, Cursor, Windsurf, or custom integration)
Setup
Choose the setup method that matches your preferred platform or tool.
Claude (Team/Enterprise)
- Visit https://claude.ai/settings/connectors.
- Click Add Custom Connection and enter the following details:
- Integration name: RudderStack
- Integration URL:
https://mcp.rudderstack.com/mcp
- Click Add.
- Authorize your RudderStack by logging in with your existing RudderStack credentials and completing the authorization process.
- Verify and test your connection.
Claude (Individual Paid)
- Visit claude.ai/settings/connectors on web or navigate to Settings > Connections on desktop.
- Click Add Custom Connection andnter the following details:
- Integration name: RudderStack
- Integration URL:
https://mcp.rudderstack.com/mcp
- Save and enable the connection.
Claude (Free)
- Open your Claude Desktop configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
- Add the RudderStack MCP server to the configuration:
{
"mcpServers": {
"rudderstack": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://mcp.rudderstack.com/mcp"]
}
}
}
- Restart Claude Desktop for the changes to take effect.
VS Code
For VS Code integration, add the following to your settings:
{
"mcp.servers": {
"rudderstack": {
"command": "npx",
"args": ["mcp-remote", "https://mcp.rudderstack.com/mcp"]
}
}
}
Cursor
Install the MCP server in Cursor
Alternatively, you can manually update your Cursor configuration:
{
"rudderstack": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://mcp.rudderstack.com/mcp"]
}
}
Windsurf
In Windsurf, navigate to Settings > MCP Servers. Then, add the following configuration:
{
"name": "rudderstack",
"command": "npx",
"args": ["mcp-remote", "https://mcp.rudderstack.com/mcp"]
}
Direct connection
For testing or using custom integrations, run the following command:
npx mcp-remote "https://mcp.rudderstack.com/mcp"
OAuth authentication flow
When you first connect to the RudderStack MCP server through any of the above methods, you will go through a secure OAuth authentication process. Here’s what happens during the setup:
- Initial connection: When you first use any MCP tool to connect to https://mcp.rudderstack.com/mcp, the server identifies that you need authentication.
- Browser redirect: Your default browser will automatically open and redirect you to RudderStack’s login page.
- RudderStack login: Sign in with your existing RudderStack credentials (email and password).
- Authorization: Review and approve the MCP integration’s requested permissions to access your RudderStack data.
- Automatic return: Once authorized, you’re redirected back and the MCP connection is established.
- Tool integration: Your MCP client (Claude, VS Code, etc.) will now have access to RudderStack tools and data.
Verify your connection
After successful setup, you should see RudderStack listed in your available tools. The tool list includes options like:
audit_logsget_destination_errorslist_connectionsget_warehouse_uploadssource_live_events, and many more
You can use these tools to get comprehensive data access and help you debug your pipeline issues.
Test your connection
Try the following prompt to confirm that the setup works as expected:
How are my RudderStack pipelines?
This prompt responds with all your sources, destinations, and connections, confirming that RudderStack AI can access your workspace data.
Use prompt templates
For effective debugging with RudderStack AI, you can use the RudderStack MCP Prompt Templates designed to help you troubleshoot common pipeline issues and get insights into your data flow.
Troubleshooting
This section contains some troubleshooting tips for the common issues you may encounter while setting up RudderStack AI.
Connection issues
If you’re having trouble connecting:
- Verify the server is accessible at https://mcp.rudderstack.com.
- Update to the latest version of npx and your MCP client.
- Try clearing stored authentication and reconnecting.
Authentication errors
- Ensure you have an active RudderStack account.
- Verify your account has the necessary permissions.
- Clear browser cookies and try again.
Common setup issues
| Issue | Solution |
|---|---|
| Connectors tab not visible in Claude | Make sure you have a paid Claude account (Pro or Team) — MCP connectors are not available on free accounts. |
| Connection not appearing | Refresh the Claude connectors and verify that you have saved the connector correctly. |
| Authorization failures | Make sure you’re logged into the correct RudderStack workspace. You can also ask Claude to switch to the desired workspace. |
| Tools not loading | You might need to re-establish the connection. Try disconnecting and reconnecting. |
Known limitations
The RudderStack MCP server is in active development. Currently, it has the following limitations:
- Some RudderStack features may not be available
- Connection stability may vary during updates
Next steps
After setting up the connector successfully, you can:
- Ask questions about your RudderStack pipeline health and configuration
- Debug data flow issues with AI assistance
- Get insights into warehouse uploads and transformations
- Explore live events from your sources
Support and feedback
This is an early access feature — if you encounter any issues, reach out to RudderStack Support.
We’d love your feedback. As you explore these AI debugging capabilities, feel free to share your experience and suggestions for improvement with the RudderStack team.
Was this page helpful? YesNo
Glad to hear it! Please tell us how we can improve.
Sorry to hear that. Please tell us how we can improve.
Questions? Contact us by Email or onSlack
- Previous
- Next
Privacy Overview
This site uses cookies to improve your experience while you navigate through the website. Out of these cookies, the cookies that are categorized as necessary are stored on your browser as they are as essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may have an effect on your browsing experience.
Always Enabled
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
This site uses cookies to improve your experience. If you want to learn more about cookies and why we use them, visit our cookie policy. We'll assume you're ok with this, but you can opt-out if you wish
Related Servers
GraphRAG
Query a hybrid graph (Neo4j) and vector (Qdrant) database for powerful semantic and graph-based document retrieval.
Amela MCP Memory Tool
A local, high-performance memory server for AI agents, built with SQLite, vector embeddings, and a knowledge graph. Packaged for npm and Docker.
PubChem MCP Server
Provides comprehensive access to PubChem's chemical information database via the PubChem PUG REST API.
Timeplus
Interact with the Timeplus real-time data platform for querying and managing data streams.
Hive MCP Server
Enables AI assistants to interact with the Hive blockchain through the Model Context Protocol.
Kuzu
Inspect schemas and execute queries on Kuzu databases.
PostgreSQL
Provides read-only access to PostgreSQL databases, enabling schema inspection and query execution.
Postgres MCP Server
Provides secure database access to PostgreSQL using the Kysely ORM.
Tabular MCP Server
An MCP server for local Tabular Models like PowerBI. It allows LLM clients to debug, analyze, and compose DAX queries by connecting to a local Tabular model instance.
Chronos
Interact with the Stellar blockchain to manage wallets, list tokens, query balances, and transfer funds.