NewsletterSubmit a tool

Official · by elastic

Elasticsearch MCP server

The Elasticsearch MCP server lets an AI assistant list indices, inspect field mappings, run searches and ES|QL queries, and check shard information.

  • Apache-2.0 licence
  • ★ 719 on GitHub
  • Verified against its source 19 Oct 2026
View source

What the Elasticsearch MCP server does

Through natural language conversations, an assistant can use the server's tools to work with Elasticsearch data. The tools include listing available indices, getting field mappings for a specific index, performing searches using query DSL, executing ES|QL queries, and getting shard information for all or specific indices. The quoted setup refers to connecting with a cluster URL and an API key.

Tools it gives your assistant

5 tools or toolsets, as its documentation lists them.

  • list_indices Lists available indices
  • get_mappings Gets field mappings for an index
  • search Runs a search using query DSL
  • esql Runs an ES QL query
  • get_shards Gets shard information for all or specific indices

How to install it

What you need

In the elastic documentation's own words:

  • “Replace with your Elasticsearch cluster URL and with your API key”
  • “If you're using Claude Desktop (free edition) which only supports the stdio protocol, use mcp-proxy to bridge stdio to streamable-HTTP”

Commands and settings change between versions. Check the official instructions before you install.