Kibana

You can use this MCP server to let AI assistants interact with Kibana dashboards, visualizations, and Elasticsearch data through a standardized interface. It exposes read-only Kibana resources and practical tools, and supports both local stdio and HTTP/SSE transports for flexible integration.

How to use

Connect an MCP client to the Kibana MCP Server over HTTP/SSE or stdio to list dashboards, fetch details, export dashboards, or run searches against Elasticsearch data. You can use Claude Code, Amazon Q Developer, or Claude Desktop in stdio mode to interact with the server. In HTTP mode, your client can stream data via SSE for real-time responses.

How to install

Prerequisites: you need Node.js and npm installed on your machine. You can also run the server with Docker or Podman for containerized deployments.

1. Install Node.js and npm if you don’t already have them. Follow your platform’s installation method for Node.js.

2. Install dependencies and run in development mode.

# Install dependencies
npm install

# Run in development stdio mode
npm run dev

# Run in development HTTP mode (for testing)
npm run dev:http

3. Build and run for production. The final startup command shown here is the stdio start command.

# Build the production artifacts
npm run build

# Start in stdio mode
npm start

# Start in HTTP mode
npm start:http

Configuration and environment

Configure the server using environment variables. You can provide Kibana connection details and transport preferences to tailor how the MCP server authenticates and communicates with Kibana.

# Kibana configuration (required)
KIBANA_URL=https://your-kibana-instance.com
KIBANA_API_KEY=your_api_key_here

# Alternative: username/password authentication
# KIBANA_USERNAME=your_username
# KIBANA_PASSWORD=your_password

# Server configuration
MCP_TRANSPORT=http           # or stdio
HTTP_PORT=3000               # Port for HTTP server
LOG_LEVEL=info               # debug, info, warn, error

Security and maintenance

The server runs with a non-root user, uses a minimal Node.js base image, and supports API key or basic authentication. It respects Kibana RBAC settings to ensure you only access allowed dashboards and data.

Troubleshooting

If you can’t reach Kibana or the MCP server, verify network connectivity and Kibana authentication. Check logs for startup errors and verify the MCP transport mode you are using.

# Health check
curl http://localhost:3000/health

# Server info
curl http://localhost:3000/info

Docker deployment

You can run the Kibana MCP Server in a container for easy deployment and isolation.

docker build -t kibana-mcp:latest .

# Run container with Kibana credentials
docker run -d \
  --name kibana-mcp \
  -p 3000:3000 \
  -e KIBANA_URL=https://your-kibana.com \
  -e KIBANA_API_KEY=your-api-key \
  kibana-mcp:latest

Development

Project structure and guidelines help you extend the server with new tools or resources. Follow the existing patterns to add new MCP endpoints and tests.

Available tools

list_dashboards

List dashboards with optional search and pagination.

get_dashboard

Get detailed information about a specific dashboard.

export_dashboard

Export a dashboard with all dependencies.

search_logs

Query Elasticsearch data through Kibana with a custom query.

list_visualizations

List available visualizations.

get_visualization

Get details for a specific visualization.

list_data_views

List available Kibana data views.