Lets Gemini CLI explore, search and query your Elasticsearch data with ES|QL, through Elastic's Agent Builder MCP server.
by elastic· DatabasesUnclaimed · Yours? Claim it
4 images
About
Elastic's official extension for Gemini CLI connects the agent to your Elasticsearch data. It brings in the Elastic Agent Builder MCP server, so Gemini CLI can list your indices, read their mappings, find documents and run ES|QL queries from plain-language questions. It also adds /elastic commands that call those tools directly. What you can do with it - List your indices, aliases and data streams, or find the relevant ones from a description - See an index's mapping and the fields it holds - Search documents with full-text or structured queries, and fetch one by its ID - Turn a question into an ES|QL query and run it, with the results as a table - Count, aggregate and summarise data, such as the top error codes in your logs this week - Run a tool directly with /elastic:list-indices, /elastic:explore-indices, /elastic:get-mapping, /elastic:search-dsl and /elastic:search-esql How it connects Install it with gemini extensions install https://github.com/elastic/gemini-cli-elasticsearch. The extension reaches Elastic Agent Builder's MCP server through mcp-remote, using two variables you set in your shell: ELASTIC_MCP_URL, copied from Kibana (Agents > View all tools > Manage MCP > Copy MCP Server URL), and ELASTIC_API_KEY, an encoded Elasticsearch API key. The MCP server needs Elastic Serverless, or Elasticsearch and Kibana 9.3 or later, with Agent Builder enabled. Agent Skills The extension includes a skills.js script that installs Elastic's Agent Skills for Elasticsearch, Kibana, Observability, Security and Elastic Cloud from the elastic/agent-skills repository. The skills call Elasticsearch's REST API directly, so they need ELASTIC_URL and your API key but not Kibana. Where a skill overlaps with the MCP tools, you can turn the server off with /mcp disable elastic-agent-builder. Good to know - The extension is experimental. - Elastic suggests an API key with read privileges only, such as feature_agentBuilder.read. - The tools come from Agent Builder, so tools you add or change in Kibana show up in Gemini CLI too.
Once it’s added
“Show me all my Elasticsearch indices.”
“Search for documents about 'error' in my logs.”
“What fields are in my user-data index?”
“Show me the top 10 error codes from my logs this week.”
Before you add it
Reads only
Indices, aliases and data streams · Index mappings and fields · Documents, by search or by ID · ES|QL query results, as a table
From people who use it
Step by step
Plugins people install alongside Elasticsearch extension for Gemini CLI.
No reviews yet. Used Elasticsearch extension for Gemini CLI? Say how it went.
Write a reviewRuns on your machine.
gemini extensions install https://github.com/elastic/gemini-cli-elasticsearch
Scanned and passed · checked 6 h ago
Plugins people install alongside Elasticsearch extension for Gemini CLI.
Equips your coding agent with a database architect, a SQL specialist and a PostgreSQL table-design skill.
Let your agent explore schemas and query your databases, from Postgres and MySQL to BigQuery, AlloyDB and Spanner.
Let your agent work with your Firebase projects: Firestore data, Auth users, security rules, Crashlytics and deploys.
Supabase's own MCP server: your agent manages tables, runs SQL and migrations, checks logs and deploys Edge Functions in your projects.
Nobody has shared how they use Elasticsearch extension for Gemini CLI yet.
“How many customers are based in the USA in the kibana_sample_data_ecommerce index?”
“Find the top 5 product categories by total sales revenue in the kibana_sample_data_ecommerce index.”
Nothing new reaches your agent until a person has looked at it.
Scanned on every change, and again when our rules improve.
Tell us and a person on the trust team will look at it. Or email [email protected].
Security scan
Passed
Safe to install in every agent it supports. Extensions.
