Kibana

Community

MCP server to access and manage Kibana instances with natural language

How to install ↓ Compare GitHub source repository ↗

Category: Dev Tools #243 of 438
Stars: ★ 84 Popular
Transport: stdio (local process) · Streamable HTTP
Runtime: Node.js 18+
Credentials: API key / credential required
License: Apache-2.0
Last commit: 1 day ago
Tools: 16

51FMRS · D

This server is based on official Kibana API, offering comprehensive Kibana management capabilities, community-maintained, suitable for development environments and personal projects.

Strongest · Documentation 14/20 Weakest · Reliability 6/20

Reliability

6/20

Security and permissions

8/20

Maintenance

10/20

Documentation

14/20

Setup experience

13/20

Why each score

Reliability 6/20: Only repo metadata, manifest, and README are available for static review; actual source, tests, and CI workflows cannot be inspected, and nothing is executed. The manifest claims stdio/HTTP dual transport and dynamic API discovery, but there is no verifiable proof. Per static calibration rules, scores are reduced; a conservative 6 is given.
Security and permissions 8/20: The manifest marks KIBANA_API_KEY and KIBANA_PASSWORD as secrets, indicating credential awareness; but no source is available to verify credential storage and transmission. README supports cookie auth, NODE_TLS_REJECT_UNAUTHORIZED=0, and exposes a broad execute_kb_api tool; HTTP mode lacks access-control discussion. Delete tools exist without explicit user confirmation. Deductions are made for missing evidence, but no red-line malware/credential theft is evident, so the score is 8.
Maintenance 10/20: The repository is not archived, has an Apache-2.0 license, and is at version 0.9.0, suggesting basic maintenance. However, there is no evidence of commit history, release cadence, dependency updates, CI, or a security response channel. Because evidence is lacking, a score of 10 reflects only the visible information.
Documentation 14/20: The README covers installation, authentication methods, configuration variables, tool lists, resource URIs, example queries, troubleshooting, and debugging. It is generally usable, but has minor inconsistencies (e.g., NODE_TLS_REJECT_UNAUTHORIZED example meaning vs. description), and lacks full parameter details, limitations, and cost notes. Score is 14.
Setup experience 13/20: Installation paths are clear: npm global, npx, and build-from-source are documented; a Claude Desktop configuration example and environment variable list are provided, making mainstream client setup simple. However, there are no tests or CI evidence to confirm end-to-end functionality; static calibration caps this at 15, so a conservative 13 is given.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

Teams using Kibana for log analysis, monitoring, and visualization
Users who want to manage Kibana resources within AI assistants like Claude
Developers integrating Elastic Stack workflows

Not for

Not for direct Elasticsearch data operations (a companion Elasticsearch MCP server exists)
Not an official Elastic product; community-maintained, not for mission-critical production use without support

Required permissions

Requires access to the Kibana server (KIBANA_URL)
Needs authentication credentials (API Key, username/password, or cookies) to perform operations
In HTTP mode, requires opening a network port (default 3000)

Risks and side effects

Misconfiguration could lead to unauthorized access or data leakage
Executing API requests may modify or delete saved objects, risking data loss
SSL verification disablement (NODE_TLS_REJECT_UNAUTHORIZED=0) may expose to man-in-the-middle attacks
Community-maintained, no official support, updates may carry risks

Setup

Before you start

Runtime：Node.js 18+

KIBANA_URL required Required Kibana instance URL, e.g. http://localhost:5601.

KIBANA_API_KEY optionalsecret Base64-encoded Kibana API key (preferred auth), generated in Kibana Stack Management.

KIBANA_PASSWORD optionalsecret Password for basic auth, used with KIBANA_USERNAME.

KIBANA_COOKIES optionalsecret Session cookie string (e.g. sid=xxx); third-priority auth method.

Other optional settings (8)

KIBANA_USERNAME optional Username for basic auth, used with KIBANA_PASSWORD if not using an API key.

KIBANA_DEFAULT_SPACE optional Default Kibana space, defaults to 'default'.

KIBANA_CA_CERT optional Path to a custom CA certificate for SSL/TLS validation.

KIBANA_TIMEOUT optional Request timeout in milliseconds, defaults to 30000.

MCP_TRANSPORT optional Transport mode selection: stdio or http, defaults to stdio.

MCP_HTTP_PORT optional HTTP mode server port, defaults to 3000.

MCP_HTTP_HOST optional HTTP mode server host, defaults to localhost.

NODE_TLS_REJECT_UNAUTHORIZED optional Set to 0 to disable SSL validation (debugging only, use with caution).

Install globally: npm install -g @tocharianou/mcp-server-kibana, or run directly with npx: npx @tocharianou/mcp-server-kibana. Set the environment variable KIBANA_URL to point to your Kibana instance. Optional authentication: KIBANA_API_KEY or KIBANA_USERNAME/KIBANA_PASSWORD. For HTTP mode, set MCP_TRANSPORT=http.

claude_desktop_config.json

{
  "mcpServers": {
    "kibana": {
      "command": "npx",
      "args": [
        "@tocharianou/mcp-server-kibana"
      ],
      "env": {
        "KIBANA_URL": "http://your-kibana-server:5601",
        "KIBANA_API_KEY": "your-api-key",
        "KIBANA_DEFAULT_SPACE": "default"
      }
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json

{
  "servers": {
    "kibana": {
      "command": "npx",
      "args": [
        "@tocharianou/mcp-server-kibana"
      ],
      "env": {
        "KIBANA_URL": "http://your-kibana-server:5601",
        "KIBANA_API_KEY": "your-api-key",
        "KIBANA_DEFAULT_SPACE": "default"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal

claude mcp add kibana -e KIBANA_URL=http://your-kibana-server:5601 -e KIBANA_API_KEY=your-api-key -e KIBANA_DEFAULT_SPACE=default -- npx @tocharianou/mcp-server-kibana

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Confirm that tools like get_status and vl_search_saved_objects appear in your client's tool list, then ask "What is the status of my Kibana server?" — a returned status proves the connection works.

Troubleshooting

Verify Kibana URL is accessible and network is reachable
Check authentication credentials for correctness and sufficient permissions
For SSL issues, try setting NODE_TLS_REJECT_UNAUTHORIZED=0 (use with caution)
If 'import: command not found', update to the latest version
Use MCP Inspector for debugging: npm run inspector

Things to try

Once connected, you can ask your AI assistant things like:

What is the status of my Kibana server?
List all available Kibana spaces
Find visualizations containing 'nginx' in the title
Check health of dashboard 'overview'

Tools 16

get_status read-only

Get Kibana server status

execute_kb_api writes

Execute custom Kibana API requests

get_available_spaces read-only

List available Kibana spaces

search_kibana_api_paths read-only

Search API endpoints

list_all_kibana_api_paths read-only

List all API endpoints

get_kibana_api_detail read-only

Get API endpoint details

vl_search_saved_objects read-only

Search saved objects (universal)

vl_get_saved_object read-only

Get single saved object

Show 8 more tools

vl_create_saved_object writes

Create new saved object

vl_update_saved_object writes

Update single saved object

vl_bulk_update_saved_objects writes

Bulk update operations

vl_bulk_delete_saved_objects destructive

Bulk delete operations

analyze_object_dependencies read-only

Analyze saved object dependencies

analyze_deletion_impact read-only

Check impact before deletion

check_dashboard_health read-only

Dashboard health check

scan_all_dashboards_health read-only

Batch health scanning

Use cases

Get Kibana server status and available spaces

Search, create, update, and delete saved objects like dashboards and visualizations

Analyze dependencies and deletion impact of saved objects

Check dashboard health

Interact with Kibana API directly from MCP clients

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

Kibana MCP Server allows any MCP-compatible client (such as Claude Desktop) to access your Kibana instance via natural language or programmatic requests. It features dynamic API discovery based on the official Kibana OpenAPI specification, providing capabilities including status checks, saved object management, dependency analysis, and health checks. Supports multiple authentication methods (API Key, Basic Auth, Cookie) and dual transport modes (stdio and HTTP).

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 62.9k · Tools 2 Compare with this →

NPM Sentinel MCP Server 79 · B

AI-powered NPM package analysis MCP server

★ 18 · Tools 19 Compare with this →

MartinLoop MCP Server 79 · B

An independent control layer that gives AI coding agents budgets, brakes, and receipts

★ 410 Compare with this →

GitHub MCP Server 76 · B

GitHub's official MCP server for managing repos, issues, PRs, and workflows via natural language

★ 33.5k · Tools 15 Compare with this →

Compare with other MCP servers

Source revision 1fe0eb9a54c7 Data synced 2026-10-12 Read the FMRS scoring method