Kibana — MCP server for Claude Code: what it does, how to install
An MCP server that connects Kibana to your AI client, so you can check status, search saved objects and call the Kibana API from chat.
claude mcp add --env KIBANA_URL=<KIBANA_URL> --env KIBANA_API_KEY=<KIBANA_API_KEY> --env KIBANA_USERNAME=<KIBANA_USERNAME> --env KIBANA_PASSWORD=<KIBANA_PASSWORD> --env MCP_TRANSPORT=<MCP_TRANSPORT> --transport stdio kibana -- npx -y @tocharianou/mcp-server-kibanaWhat Kibana does
The server passes requests to your Kibana instance: server status, available spaces, search across saved objects (dashboard, visualization, index-pattern, lens, map and more), bulk updates and deletes. A separate tool searches Elasticsearch documents through Kibana's search API, limited by your index privileges. For anything else there is execute_kb_api, which sends a custom request to the Kibana API.
Two transports: stdio for local runs and HTTP for remote access. Auth is an API key, basic auth or cookies. Start HTTP mode on 0.0.0.0 without MCP_BEARER_TOKEN and the server logs a warning at startup: anyone who reaches the port uses your Kibana credentials. Behind a reverse proxy that forwards its public Host to a localhost-bound server, list that hostname in MCP_ALLOWED_HOSTS.
Only KIBANA_URL is required. Request timeout defaults to 60000 ms, the space to default. The project is community-maintained and not an official Elastic product.
What it helps with
- check Kibana status without opening a browser
- find a saved object by name and type
- bulk update or delete dashboards
- see what breaks before deleting an object
- run a custom Kibana API request
Example prompts for the agent
- Show Kibana status and the available spaces
- Find saved objects of type dashboard with revenue in the title
- Run a Kibana API request to list index patterns
How to install Kibana
Platform CLI
claude mcp add --env KIBANA_URL=<KIBANA_URL> --env KIBANA_API_KEY=<KIBANA_API_KEY> --env KIBANA_USERNAME=<KIBANA_USERNAME> --env KIBANA_PASSWORD=<KIBANA_PASSWORD> --env MCP_TRANSPORT=<MCP_TRANSPORT> --transport stdio kibana -- npx -y @tocharianou/mcp-server-kibanaReplace <…> with your own values. 2 more environment variables are in the server README
Platform CLI
codex mcp add kibana --env KIBANA_URL=<KIBANA_URL> --env KIBANA_API_KEY=<KIBANA_API_KEY> --env KIBANA_USERNAME=<KIBANA_USERNAME> --env KIBANA_PASSWORD=<KIBANA_PASSWORD> --env MCP_TRANSPORT=<MCP_TRANSPORT> -- npx -y @tocharianou/mcp-server-kibanaReplace <…> with your own values. 2 more environment variables are in the server README
Config file
{
"mcpServers": {
"kibana": {
"command": "npx",
"args": [
"-y",
"@tocharianou/mcp-server-kibana"
],
"env": {
"KIBANA_URL": "<KIBANA_URL>",
"KIBANA_API_KEY": "<KIBANA_API_KEY>",
"KIBANA_USERNAME": "<KIBANA_USERNAME>",
"KIBANA_PASSWORD": "<KIBANA_PASSWORD>",
"MCP_TRANSPORT": "<MCP_TRANSPORT>"
}
}
}
}Replace <…> with your own values. 2 more environment variables are in the server README
Config file
{
"servers": {
"kibana": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@tocharianou/mcp-server-kibana"
],
"env": {
"KIBANA_URL": "<KIBANA_URL>",
"KIBANA_API_KEY": "<KIBANA_API_KEY>",
"KIBANA_USERNAME": "<KIBANA_USERNAME>",
"KIBANA_PASSWORD": "<KIBANA_PASSWORD>",
"MCP_TRANSPORT": "<MCP_TRANSPORT>"
}
}
}
}Replace <…> with your own values. 2 more environment variables are in the server README
Platform CLI
gemini mcp add -e KIBANA_URL=<KIBANA_URL> -e KIBANA_API_KEY=<KIBANA_API_KEY> -e KIBANA_USERNAME=<KIBANA_USERNAME> -e KIBANA_PASSWORD=<KIBANA_PASSWORD> -e MCP_TRANSPORT=<MCP_TRANSPORT> kibana npx -y @tocharianou/mcp-server-kibanaReplace <…> with your own values. 2 more environment variables are in the server README
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Kibana
How do I install the Kibana MCP server in Claude Code?
Add it with `claude mcp add kibana -e KIBANA_URL=<KIBANA_URL> -e KIBANA_API_KEY=<KIBANA_API_KEY> -e KIBANA_PASSWORD=<KIBANA_PASSWORD> -- npx -y @tocharianou/mcp-server-kibana. Alternatively run claude mcp add --env KIBANA_URL=<KIBANA_URL> --env KIBANA_API_KEY=<KIBANA_API_KEY> --env KIBANA_USERNAME=<KIBANA_USERNAME> --env KIBANA_PASSWORD=<KIBANA_PASSWORD> --env MCP_TRANSPORT=<MCP_TRANSPORT> --transport stdio kibana -- npx -y @tocharianou/mcp-server-kibana`.
What does the Kibana MCP server do?
It gives an AI client tools for Kibana: server status, spaces, search and CRUD on saved objects, dependency analysis, and custom Kibana API requests.
Which variables does it need?
KIBANA_URL is required. Pick one auth method: KIBANA_API_KEY, KIBANA_USERNAME with KIBANA_PASSWORD, or KIBANA_COOKIES. The rest are optional.
