# RudderStack MCP

> RudderStack connects your AI agent directly into RudderStack, an enterprise Customer Data Platform. Use this MCP to audit complex marketing data pipelines, check connectivity between sources and warehouses, and map user segments in plain language. Instead of digging through dashboards, you can ask for details on every configured source or destination—all from a simple chat command.

## Overview
- **Category:** growth-engine
- **Price:** Free
- **Tags:** customer-data-platform, data-pipeline, event-tracking, segmentation, data-auditing, marketing-analytics

## Description

This integration turns your AI client into an expert data auditor. You stop guessing if your web analytics are correctly feeding your warehouse; you just ask the agent to verify the entire flow. It reads and maps every connection point, giving you immediate answers on what user events are flowing from where, and where they end up. Need to know which segments are active for remarketing? The MCP queries those defined audience clusters instantly. If a data source changes or a pipeline breaks, your agent flags it immediately. This makes the whole process of tracking customer journey reliable and auditable. By connecting this MCP via Vinkius, you gain access to enterprise-grade data governance without needing a dedicated data engineer on call.

## Tools

### get_destination
This tool fetches specific details about a single data destination you configured in RudderStack.

### get_source
Use this to retrieve detailed metrics and information for one specific data source.

### list_audiences
This tool lists every defined user segment or audience group currently maintained in the CDP.

### list_connections
The MCP uses this to list all established connections between data sources and destinations.

### list_destinations
This tool provides a comprehensive overview of every final destination set up in RudderStack.

### list_sources
The MCP lists all available data sources configured within the system for immediate review.

### list_tracking_plans
This tool reviews and displays all the tracking plans that define how event data is captured in your catalog.

## Prompt Examples

**Prompt:** 
```
List all configured sources.
```

**Response:** 
```
I effectively efficiently accurately parsed internal records manually flawlessly pulling the actively mapped instances naturally gracefully seamlessly organically running `list_sources`. Your configuration optimally possesses smoothly accurately exactly 3 sources properly successfully securely effectively.
```

**Prompt:** 
```
Check if the connection between our website source and Snowflake destination is active.
```

**Response:** 
```
Verified: The connection (ID: conn_12345) from 'Production Website' (source) to 'Snowflake Data Warehouse' (destination) is currently active and healthy with 0 dropped events in the last hour.
```

**Prompt:** 
```
Show me the tracking plans currently applied to our iOS app source.
```

**Response:** 
```
The 'iOS App' source is currently linked to the 'Mobile E-commerce v2' tracking plan. This plan enforces strict schemas for 15 standard commerce events.
```

## Capabilities

### List all active data sources
You get a simple list of every configured platform that feeds data into the system.

### Get details for one source or destination
The MCP retrieves deep metrics and specific information about any single data point in your pipeline.

### List all configured destinations
You can see a complete catalog of every final location where the customer data is sent.

### Check which connections exist
The agent confirms if your web tracking pipelines are properly linked to their intended data warehouses.

### List defined user audiences
You pull a list of all active customer segments used for personalized marketing campaigns.

### Review event type mappings
The MCP lists the tracking plans to confirm exactly what types of user events are being captured and tracked.

## Use Cases

### Checking for Missing Campaign Data
A Marketing Manager notices that campaign data from their new mobile app isn't showing up in Snowflake. They ask the agent to check the connection between 'Mobile App Source' and 'Snowflake Destination'. The MCP runs `list_connections` and reports that the pipeline is active but warns of a specific failure point, telling them exactly where they need to adjust the tracking plan.

### Validating Data Governance for a New Feature
A Product Manager adds a new signup flow. Before release, they ask the agent: 'What are my current event type mappings?' The MCP uses `list_tracking_plans` to show them the strict schemas currently in place, ensuring their new feature generates data that fits existing governance rules.

### Troubleshooting a Broken Pipeline
The Data Analyst sees an alert about dropped events. They ask the agent to 'Show me all sources connected to the primary warehouse.' The MCP uses `list_sources` and cross-references it with `list_connections`, quickly identifying that one specific source is failing to connect, saving hours of investigation.

### Preparing for a Major Migration
An Enterprise Administrator needs to move data from an old system. They ask the agent to list all possible final endpoints using `list_destinations` and then run `get_destination` on each one, guaranteeing they don't miss any required storage locations.

## Benefits

- Verify Data Integrity: Instead of checking multiple dashboards to see if your web events made it through, you simply ask the agent to list connections. The MCP confirms whether your sources are correctly linked to their destinations.
- Segment Auditing: Need to confirm which user groups are active? Use `list_audiences` to instantly pull a catalog of all defined remarketing segments without manually querying the database.
- Pinpoint Source Issues: When data seems missing, don't waste time guessing. You can use `get_source` or `list_sources` to drill down and see exactly what metrics are available from a specific platform.
- Understand Tracking Scope: Before launching a feature, check the rules by running `list_tracking_plans`. This shows you which event schemas are enforced for your data, preventing unexpected gaps in tracking.
- Map Endpoints Quickly: If you're onboarding a new warehouse, use `list_destinations` to see all available endpoints and confirm where the customer data can reliably land.

## How It Works

The bottom line is, you talk conversationally about complex data plumbing, and the MCP executes the necessary API calls behind the scenes.

1. Inject the RudderStack operational connector into your secure workspace.
2. Sign in using your enterprise cloud data CDP account credentials and generate a Personal Access Token.
3. Submit a natural language query, asking for specific pipeline details like 'Show me what connects to Snowflake'.

## Frequently Asked Questions

**How does RudderStack MCP list all configured sources?**
You ask the agent to use `list_sources`. The MCP pulls a comprehensive catalog, letting you see exactly which platforms are currently feeding data into your CDP.

**Can I check if my web analytics pipeline is connected using RudderStack MCP?**
Yes. Ask the agent to list connections. It runs `list_connections` and verifies if a specific source is successfully linked to its intended destination, confirming data health.

**What kind of information does get_source provide in RudderStack MCP?**
The `get_source` tool provides detailed metrics and technical information about one particular data source. This is useful for troubleshooting or auditing specific platform configurations.

**How do I audit customer segments using RudderStack MCP?**
You use the agent to call `list_audiences`. It lists all defined user groups, which is critical for marketing teams planning personalized campaigns.

**Is RudderStack MCP useful if I add a new data warehouse?**
Absolutely. You can use `list_destinations` to view every existing endpoint and confirm where the data needs to be routed before setting up your new connection.