LlamaCloud MCP Server

The LlamaCloud MCP Server connects your LlamaCloud knowledge base to Ejento as a tool. Once set up, Ejento can search and retrieve content from your indexed documents in real time — no hardcoded data, no retraining. Just point it at your index and ask questions in natural language.

Prerequisites

Before deploying this MCP server, make sure you have the following ready:

A LlamaCloud account at cloud.llamaindex.ai

A LlamaCloud index already created with at least one document uploaded and indexed

A LlamaCloud API key from https://cloud.llamaindex.ai/api-key

Your project name (visible in the LlamaCloud UI, defaults to Default)

Note: The index must exist and contain documents before the server starts. The server connects to an existing index — it does not create one.

Configuration & Connecting With Ejento

Set the following environment variables in the web app.

Variable	Required	Description
`LLAMA_CLOUD_API_KEY`	✅	Your LlamaCloud API key
`LLAMA_CLOUD_INDEX_NAME`	✅	The name of your LlamaCloud index
`LLAMA_CLOUD_PROJECT_NAME`	✅	The project the index belongs to (usually `Default`)
`LLAMA_CLOUD_ORG_ID`	optional	Organization ID — only needed if you have multiple orgs

Once configured, you can start using the connector on Ejento

Available Tools

`query_index`

The primary tool. Accepts a natural language question and returns relevant content retrieved from the index. Use this for any question that could be answered from your indexed documents.

Example: "What are the refund policies described in the document?"

`list_documents`

Returns a list of all documents currently in the index, including their names and IDs. Useful for confirming what has been indexed before querying.

Example: "What documents are in the index?"

`insert_document`

Adds a new piece of text content to the index at runtime. Requires a text body and a unique doc_id.

Example: "Add this new policy update to the index."

`delete_document`

Removes a document from the index by its ID. Use list_documents first to confirm the correct ID before deleting.

Example: "Delete the document with ID doc-001 from the index."

LlamaCloud Connector

LlamaCloud MCP Server#

Prerequisites#

Configuration & Connecting With Ejento#

Available Tools#

query_index#

list_documents#

insert_document#

delete_document#