What is Humio Mcp

humio-mcp is a Model Context Protocol (MCP) server implementation designed for integrating with Humio’s logging and analytics platform, facilitating seamless data flow between applications and Humio’s features.

Use cases

Use cases for humio-mcp include monitoring application logs in real-time, analyzing user behavior through logs, and troubleshooting issues by leveraging Humio’s powerful analytics tools.

How to use

To use humio-mcp, clone the repository, install the necessary dependencies, configure your Humio API token in a .env file, and build the server. After building, you can inspect the server using the MCP Inspector and connect it to your client application.

Key features

Key features of humio-mcp include integration with Humio’s logging capabilities, support for custom queries through a configuration file, and the ability to inspect the server using the MCP Inspector.

Where to use

humio-mcp can be used in various fields where logging and analytics are crucial, such as software development, IT operations, and data analysis, providing insights into application performance and behavior.

Humio MCP Server

A Model Context Protocol (MCP) server implementation for Humio, enabling integration with Humio’s logging and analytics platform.

Description

This project implements a server that bridges the Model Context Protocol with Humio’s logging capabilities. It allows for seamless integration between applications and Humio’s powerful logging and analytics features.

Prerequisites

• Humio token

Installation

1. Clone the repository:

git clone [repository-url]
cd humio-mcp

2. Install dependencies:

npm install

or

pnpm install

or

yarn install

3. Create a .env file in the root directory with your Humio configuration (this is only useful for dev/testing as the MCP client will provide the env variables anyway):

HUMIO_API_TOKEN=your_api_token
HUMIO_REQUEST_TIMEOUT_MS=30000

4. Create config from the example config file

cp humio-query-config.example.json humio-query-config.json

Usage

Build the server:

npm run build

or

pnpm run build

or

yarn run build

Inspection

After building, you can inspect the server using the MCP Inspector:

npx @modelcontextprotocol/inspector node dist/index.js

Connect to the client. Sample config

Configuration File Format

The humio-query-config.json file defines the available queries for the server. Each entry in the array should follow the structure below:

Variable Format

Each item in the variables array can be one of the following:

Scalar Variable

Enum Variable

Example

[
  {
    "name": "criticalErrors",
    "description": "Finds critical errors for the banana team grouped by message and stack trace.",
    "query": "k8s.labels.team = banana | (severity = crit ) | groupBy([message, stack_trace])",
    "fields": [
      "message",
      "stack_trace",
      "_count"
    ],
    "outputTemplate": "Error \"{{message}}\" occurred in total: {{_count}} times. The Stack trace is \n---{{stack_trace}}\n---\n\n",
    "joinString": "\n"
  },
  {
    "name": "projectOperationsByState",
    "description": "Groups operations by state for a given project in the banana team.",
    "query": "k8s.labels.team = banana | projectKey=\"{{projectKey}}\" | groupBy([fields.state])",
    "fields": [
      "fields.state",
      "_count"
    ],
    "variables": [
      {
        "name": "projectKey",
        "description": "The project key to filter by.",
        "type": "string",
        "required": true
      }
    ],
    "outputTemplate": "state: {{fields.state}} - {{_count}} operations",
    "joinString": "\n"
  }
]

⚠️ Danger: Sensitive Information

Do NOT include any sensitive information in the fields field of your configuration.

The contents of the fields field are returned to the client and may be passed to a Large Language Model (LLM). Any sensitive data (such as API keys, passwords, personal information, or confidential business data) included here could be exposed.

Ensure that only non-sensitive, safe-to-share information is specified in the fields field of your config files.

⚠️ Warning: Timeout and Partial Results

If the configured timeout (HUMIO_REQUEST_TIMEOUT_MS) is reached during a query, the server will return the events collected up to that point. This means the results may be incomplete and might not capture all relevant events.

To reduce the risk of missing data, consider adjusting the timeout value according to your expected query size and network conditions.

Project Structure

humio-mcp/
├── src/
│   ├── humio/                  # Humio-specific implementations
│   ├── utils/                  # Utility functions
│   └── index.ts                # Main server entry point
├── dist/                       # Compiled JavaScript files
├── humio-query-config.json     # Configurations for available quries
├── package.json                # Project dependencies and scripts
└── tsconfig.json               # TypeScript configuration