LlamaCloud
Connect to and manage data indexes on the LlamaCloud platform.
LlamaCloud MCP Server
A MCP server connecting to multiple managed indexes on LlamaCloud
This is a TypeScript-based MCP server that creates multiple tools, each connected to a specific managed index on LlamaCloud. Each tool is defined through command-line arguments.
Features
Tools
- Creates a separate tool for each index you define
- Each tool provides a
queryparameter to search its specific index - Auto-generates tool names like
get_information_index_namebased on index names
Installation
To use with your MCP Client (e.g. Claude Desktop, Windsurf or Cursor), add the following config to your MCP client config:
The LLAMA_CLOUD_PROJECT_NAME environment variable is optional and defaults to Default if not set.
{
"mcpServers": {
"llamacloud": {
"command": "npx",
"args": [
"-y",
"@llamaindex/mcp-server-llamacloud",
"--index",
"10k-SEC-Tesla",
"--description",
"10k SEC documents from 2023 for Tesla",
"--topK",
"5",
"--index",
"10k-SEC-Apple",
"--description",
"10k SEC documents from 2023 for Apple"
],
"env": {
"LLAMA_CLOUD_API_KEY": "<YOUR_API_KEY>"
}
}
}
}
For Claude, the MCP config can be found at:
- On MacOS:
~/Library/Application Support/Claude/claude_desktop_config.json - On Windows:
%APPDATA%/Claude/claude_desktop_config.json
Tool Definition Format
In the args array of the MCP config, you can define multiple tools by providing pairs of --index and --description arguments. Each pair defines a new tool. You can also optionally specify --topK to limit the number of results.
For example:
--index "10k-SEC-Tesla" --description "10k SEC documents from 2023 for Tesla" --topK 5
Adds a tool for the 10k-SEC-Tesla LlamaCloud index to the MCP server. In this example, it's configured to return the top 5 results.
Development
Install dependencies:
npm install
Build the server:
npm run build
For development with auto-rebuild:
npm run watch
To use the development version, replace in your MCP config npx @llamaindex/mcp-server-llamacloud with node ./build/index.js.
Debugging
Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector, which is available as a package script:
npm run inspector
The Inspector will provide a URL to access debugging tools in your browser.
Related Servers
Remote MCP Server (Authless)
A remote MCP server without authentication, deployable on Cloudflare Workers.
Authorize.Net by CData
A read-only MCP server by CData for querying live Authorize.Net data.
Cloudflare MCP Server Template
A template for deploying a remote, authentication-free MCP server on Cloudflare Workers. Tools are defined directly in the source code.
EdgeOne Geo Location Service
Provides user geolocation data via Tencent EdgeOne Pages Functions, enabling large language models to access location information.
Dokploy
An AI-powered interface for managing the Dokploy infrastructure platform.
Google Ads
MCP server acting as an interface to the Google Ads, enabling programmatic access to Google Ads data and management features.
Cloudflare Remote MCP Server (Authless)
Deploy a remote, authentication-free MCP server on Cloudflare Workers or locally via npm.
MCP Spotify AI Assistant
An AI assistant that controls Spotify features like playback, playlists, and search using the Model Context Protocol (MCP).
Terrakube MCP Server
Manage Terrakube workspaces, variables, modules, and organizations.
Workday by CData
A read-only server for querying live Workday data using LLMs, powered by the CData JDBC Driver.