Skip to directory

Elastic Agent Builder MCP

Elastic (official)

Streamable HTTP7+ documented tools0

About Elastic Agent Builder MCP

Expose Elastic Agent Builder built-in and custom tools to MCP clients from a configured Kibana deployment.

What you can use it for

  • Search and query Elasticsearch data with Agent Builder tools.
  • Expose enabled workflows and project-specific custom tools to MCP clients.
  • Access tool capabilities using the connected Elastic user or API-key identity.
Connection and setup
Hosting
Remote, per Elastic deployment or Serverless project
Transport
Streamable HTTP
Endpoint
Local process; no hosted endpoint documented
Authentication
API key for Elastic Stack and Serverless, or OAuth 2.1 for Serverless projects; access follows the key’s permissions or the consenting user’s live permissions.

Requirements

  • Elastic deployment with Agent Builder and MCP access available
  • Kibana URL and chosen space, if applicable
  • API key or supported OAuth 2.1 application connection

Setup steps

  1. In Elastic Agent Builder, open the MCP server access controls and copy the URL for the intended deployment/space.
  2. Choose API-key auth for supported Elastic Stack or Serverless use, or configure OAuth 2.1 for a Serverless interactive client. Keep credentials in protected client configuration.
  3. Add the deployment-specific endpoint to the MCP client and authorize/connect.
  4. Reload the client and inspect the returned tools. Start with a read-only query; tool availability and permissions depend on the deployment and version.
Tools 7+

Crawl4AI verified the official Elastic MCP server docs and built-in tools reference. These seven stable platform.core tools are a documented subset; built-in and custom tools vary by deployment, permissions, license, and features, and some built-ins are unavailable through MCP.

  • platform.core.execute_esql

    Execute an ES|QL query against Elasticsearch data and return tabular results.

    01
  • platform.core.generate_esql

    Generate an ES|QL query from a natural-language request.

    02
  • platform.core.get_document_by_id

    Retrieve a document by ID and index name.

    03
  • platform.core.get_index_mapping

    Retrieve mappings for one or more indices.

    04
  • platform.core.index_explorer

    Find relevant indices and mappings from a natural-language query.

    05
  • platform.core.list_indices

    List indices, aliases, and data streams the current user can access.

    06
  • platform.core.search

    Search Elasticsearch data using natural language.

    07

This server’s available tools can vary by account, permissions, or configuration. The list shown is not its full runtime catalog.

Limitations and notes
  • Requires an Elastic 9.2+ deployment or supported Serverless project with Agent Builder.
  • Available tools depend on deployment version, permissions, licenses, settings, and custom configuration.
  • Some built-in tools are not exposed through MCP; no universal fixed inventory applies.
Original sources
  • official documentation

    Elastic documentation describes the deployment-scoped MCP endpoint and the built-in/custom tools it exposes; available tools depend on deployment, permissions, product version, licenses, features, and configured custom tools.

  • official tool reference

    Lists built-in tool families and identifies tools unavailable through MCP or gated by deployment/license/features.

  • official publisher profile

    Verified Elastic organization identity; publisher avatar used as fallback tier D brand asset.

Documentation checked 2026-10-29T00:00:00Z. Connection has not been runtime tested.

Similar tools