home / mcp / alkemi mcp server
A STDIO Model Context Protocol Server that lets MCP Clients query databases using plain-english questions and query exposed API endpoints abstracting data products.
Configuration
View docs{
"mcpServers": {
"alkemi-ai-alkemi-mcp": {
"url": "https://api.alkemi.cloud/mcp",
"headers": {
"MCP_NAME": "customer-data",
"PRODUCT_ID": "123",
"BEARER_TOKEN": "sk-12345"
}
}
}
}You can connect Alkemi Data to your MCP Client through a dedicated MCP Server that runs as a STDIO wrapper around the Streamable HTTP MCP Endpoint. This server stores database metadata, generates valid queries for your schemas, and executes them against your data sources so you and your teammates share consistent querying capabilities.
Choose your integration method based on your workflow. If you want to run the MCP server locally and feed data through your client via standard input/output, use the STDIO configuration with a local node process. If you prefer a remote endpoint, you can connect through the HTTP URL and let the MCP Client reach the server directly.
Prerequisites: you need Node.js and npm installed on your system.
Install dependencies and build the server during development.
npm install
```
```bash
npm run build
```
```bash
npm run watch
```
```bash
npm run inspector
```
```bash
# Optional: run the MCP Inspector to access debugging tools in your browser
npm run inspector{
"mcpServers": {
"alkemi": {
"command": "npx",
"args": [
"@alkemiai/alkemi-mcp"
],
"env": {
"BEARER_TOKEN": "sk-12345"
}
}
}
}If you prefer running the MCP server locally via STDIO, start the Node process that serves the MCP endpoint and pass your configuration through environment variables.
{
"mcpServers": {
"alkemi-data": {
"command": "node",
"args": [
"/path/to/alkemi-mcp/build/index.js"
],
"env": {
"BEARER_TOKEN": "sk-12345"
}
}
}
}Provides a URL to access debugging tools in your browser to help you inspect and troubleshoot MCP server activity.