# embeddedlayers/mcp-analytics [Health: Active]

**Category:** 🧮 Data Science Tools  
**Repository:** https://github.com/embeddedlayers/mcp-analytics  
**GitHub Stars:** 8  
**npm Downloads (last month):** 24  
**Views:** 5  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/embeddedlayers-mcp-analytics

## Description
Statistical analysis, forecasting, and ML for business data (Shopify, Stripe, WooCommerce, eBay, GA4, Search Console). Upload a CSV or connect live data sources — ask a question in Claude or Cursor, get an interactive HTML report.

## Tools
Capabilities this server exposes over MCP:

- **create_analysis**
- **build_status**
- **run_analysis**
- **discover_tools**
- **modify_analysis**
- **tools_schema**
- **datasets_upload**
- **datasets_list**
- **connectors_list**
- **connectors_query**
- **reports_view**
- **reports_list**
- **report_cards**
- **ask_library**
- **agent_advisor**
- **billing**
- **account_link**
- **about**

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "mcp-analytics": {
    "url": "https://api.mcpanalytics.ai/mcp/api-key"
  }
}
```

## Documentation

## What embeddedlayers/mcp-analytics does

embeddedlayers/mcp-analytics turns questions about business data into analysis workflows accessed through MCP. It can provide quick answers based on the full dataset, run established statistical methods, or build a custom analysis for questions specific to a business. The result is a report containing visualizations, tables, written interpretation, the method used, and the code executed.

Analyses are designed to be repeatable. After an analysis has been created, the same workflow can be run against a later file, such as the next month’s data. Reports can also be stored, listed, viewed, and queried as a library.

## How it works

The server is exposed through a Streamable HTTP JSON-RPC endpoint:

`https://api.mcpanalytics.ai/mcp/api-key`

A first request acts as a resolver. It examines the question and available data without executing an analysis or charging for that step, then identifies the suitable route and explains why. The selected route can be a straightforward answer, a library analysis, or a custom build.

Data may be uploaded as a CSV or queried through connected data sources. The listed source categories include Shopify, Stripe, WooCommerce, eBay, Google Analytics 4, and Google Search Console. The service is intended for business and statistical questions about data, rather than open-ended assistant tasks.

## Setup and configuration

Create an account at `mcpanalytics.ai`, then copy an API key from the account settings. Configure an MCP client with the endpoint and send the key in the `X-API-Key` header:

```json
{
  "mcpServers": {
    "mcpanalytics": {
      "url": "https://api.mcpanalytics.ai/mcp/api-key",
      "headers": { "X-API-Key": "mcp_your_key_here" }
    }
  }
}
```

OAuth 2.0 through Auth0 is also available from compatible AI clients. Claude Code can be configured with the HTTP transport using its `claude mcp add` command and the same endpoint and header.

The endpoint supports discovery without a key: clients can connect and list tools, but they cannot execute operations. A direct `tools/list` JSON-RPC request can be sent with `curl` to inspect the available interface.

## Tools and capabilities

The exposed tools cover the full analysis lifecycle and related account operations:

- `datasets_upload` and `datasets_list` manage uploaded datasets.
- `connectors_list` and `connectors_query` inspect and query connected sources.
- `create_analysis`, `modify_analysis`, `run_analysis`, and `build_status` create, adjust, execute, and inspect analysis jobs.
- `reports_view`, `reports_list`, and `report_cards` retrieve reports and report summaries.
- `ask_library` asks questions across the saved report library.
- `discover_tools` and `tools_schema` expose tool discovery and schemas.
- `agent_advisor` supports the analysis-routing workflow.
- `account_link`, `billing`, and `about` provide account and service information.

Example questions include comparing revenue between months, identifying customers likely to churn, and forecasting the next quarter from an uploaded file.

## Limitations and notes

The service requires an account and API key, or an available Auth0 OAuth flow, to run operations. Discovery is available without authentication, but execution is not. Uploaded data is used to answer questions; retention details are governed by the service’s privacy policy. Reports name their methods and executed code, but the server is intended for data analysis rather than general-purpose assistance.

_Full upstream README: https://allmcps.com/mcp/embeddedlayers-mcp-analytics/readme_