Elasticsearch MCP Server

What is Elasticsearch MCP Server?

A Model Context Protocol (MCP) server implementation that provides Elasticsearch interaction. It enables searching documents, analyzing indices, and managing a cluster through a set of tools, designed for use with MCP clients like Claude Desktop.

How to use Elasticsearch MCP Server?

Configure the server in your MCP client (e.g., Claude Desktop) by setting environment variables ELASTIC_HOST, ELASTIC_USERNAME, and ELASTIC_PASSWORD to point to your Elasticsearch cluster. Use the provided tools via natural language commands, such as "List all indices in the cluster" or "Show me the cluster health status."

Key features of Elasticsearch MCP Server

• list_indices: List all indices in the Elasticsearch cluster.
• get_mapping: Retrieve mapping configuration for a specific index.
• get_settings: Get settings configuration for a specific index.
• search_documents: Search documents using Elasticsearch Query DSL.
• get_cluster_health: Get health status of the cluster.
• get_cluster_stats: Get statistical information about the cluster.

Use cases of Elasticsearch MCP Server

• Query and explore indices in your Elasticsearch cluster via natural language.
• Monitor cluster health and statistics without manual API calls.
• Retrieve index mappings and settings for troubleshooting or documentation.
• Search documents using complex Query DSL without writing raw JSON.

FAQ from Elasticsearch MCP Server

What is the Elasticsearch MCP Server?

It is an MCP server that exposes Elasticsearch operations (index, document, and cluster management) as callable tools for AI assistants like Claude.

How do I connect to my Elasticsearch cluster?

Set the ELASTIC_HOST, ELASTIC_USERNAME, and ELASTIC_PASSWORD environment variables in your MCP client configuration to match your cluster’s credentials.

What tools does the server provide?

The server provides six tools: list_indices, get_mapping, get_settings, search_documents, get_cluster_health, and get_cluster_stats.

What are the prerequisites?

You need an Elasticsearch cluster (local or remote) and the MCP server running via uv (Python). A sample Docker Compose file is provided to start a 3-node cluster with Kibana.

Does the server support authentication?

Yes, basic authentication is supported via the ELASTIC_USERNAME and ELASTIC_PASSWORD environment variables. The default credentials in the provided Docker Compose are elastic / test123.