MCP Connector

Manage TheHive Security Incidents

Manage TheHive security incident response - alerts, cases, tasks, observables, and Cortex analyzers - from an AI assistant.

Works with thehivecortex

91
Spark score
out of 100
Updated 10 months ago
Version 1.0.0
Models
universal

Add to Favorites

Why it matters

Automate and streamline your security incident response by providing standardized access to TheHive's case management, alert processing, and observable tracking capabilities. Integrate seamlessly with Cortex for enhanced analysis.

Outcomes

What it gets done

01

Manage TheHive cases, alerts, and observables

02

Automate alert processing and conversion to cases

03

Track and analyze observables with similarity detection

04

Integrate with Cortex for automated analysis

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/vb-thehive-mcp-server | bash

Capabilities

Tools your agent gets

add_case_attachment

Add attachments to cases

assign_task

Assign tasks to users

bulk_delete_alerts

Delete multiple alerts at once

bulk_delete_observables

Delete multiple observables at once

bulk_merge_alerts_into_case

Merge multiple alerts into a case

bulk_update_alerts

Update multiple alerts at once

bulk_update_cases

Update multiple cases at once

bulk_update_observables

Update multiple observables at once

+7 tools

Overview

TheHive MCP Server

TheHive MCP Server gives Claude tools to manage TheHive alerts, cases, tasks, and observables, including running Cortex analyzers and responders. Use it for conversational SOC/incident-response workflows in TheHive. Review bulk operations before running them since they can affect many records at once.

What it does

TheHive MCP Server integrates TheHive's security incident response platform with MCP clients, exposing tools to manage alerts, cases, tasks, observables, and Cortex analyzer/responder jobs.

When to use - and when NOT to

Use it to triage and manage incident response work conversationally: create or update alerts, promote/merge alerts into cases, assign and track tasks, add and analyze observables, or run Cortex analyzers and responders on indicators. It requires a running TheHive instance and API key, and operates directly on live case/alert data - since bulk operations (bulk_delete_alerts, bulk_update_cases, etc.) can affect many records at once, review bulk actions before running them.

Capabilities

  • Alerts: create_alert, get_alert(s), update_alert, delete_alert, count_alerts, follow_alert/unfollow_alert, promote_alert_to_case, merge_alert_into_case, bulk_merge_alerts_into_case, bulk_update_alerts, bulk_delete_alerts, import_alert_into_case.
  • Cases: create_case, get_case(s), update_case, close_case, delete_case, merge_cases, count_cases, bulk_update_cases, plus case pages, procedures, comments, and attachments (create_case_page, create_case_procedure, add_case_attachment, download_case_attachment, delete_case_attachment).
  • Tasks: create_task, get_task(s), assign_task, start_task, complete_task, update_task, delete_task, create_task_log, bulk_update_tasks, count_tasks.
  • Observables: create_observable_in_alert/create_observable_in_case, get_observable(s), update_observable, delete_observable, share_observable/unshare_observable, get_alert_similar_observables/get_case_similar_observables, bulk_update_observables, bulk_delete_observables, count_observables.
  • Cortex integration: list_cortex_analyzers(_by_type), get_cortex_analyzer, run_observable_analyzer(s), create_cortex_analyzer_job, get_cortex_analyzer_job, get_observable_analyzer_jobs, list_cortex_responders, create_cortex_responder_action.

How to install

uvx thehive-mcp-server

Configure Claude Desktop:

{
  "mcpServers": {
    "thehive-mcp-server": {
      "command": "uvx",
      "args": ["thehive-mcp-server"],
      "env": {
        "HIVE_URL": "https://your-thehive-host:9000",
        "HIVE_API_KEY": "your-api-key"
      }
    }
  }
}

Requires a running TheHive instance and API key (HIVE_URL, HIVE_API_KEY). Built on the official thehive4py client library, installed automatically as a dependency.

Who it's for

SOC analysts and incident responders using TheHive who want an AI assistant to help triage alerts, manage cases and tasks, and run Cortex analyzers on observables.

Source README

TheHive MCP Server

A Model Context Protocol (MCP) server implementation for TheHive, enabling seamless integration with MCP clients. This project provides a standardized way to interact with TheHive's security incident response platform through the Model Context Protocol.

Implemented MCP Tools

The following MCP tools are implemented and available:

  • add_case_attachment
  • assign_task
  • bulk_delete_alerts
  • bulk_delete_observables
  • bulk_merge_alerts_into_case
  • bulk_update_alerts
  • bulk_update_cases
  • bulk_update_observables
  • bulk_update_tasks
  • close_case
  • complete_task
  • count_alerts
  • count_cases
  • count_observables
  • count_tasks
  • create_alert
  • create_alert_observable
  • create_case
  • create_case_observable
  • create_case_page
  • create_case_procedure
  • create_case_task
  • create_cortex_analyzer_job
  • create_cortex_responder_action
  • create_observable_in_alert
  • create_observable_in_case
  • create_task
  • create_task_log
  • delete_alert
  • delete_case
  • delete_case_attachment
  • delete_observable
  • delete_task
  • download_case_attachment
  • find_alert_observables
  • find_case_attachments
  • find_case_comments
  • find_case_observables
  • find_case_pages
  • find_case_procedures
  • find_case_tasks
  • find_task_logs
  • follow_alert
  • get_alert
  • get_alert_similar_observables
  • get_alerts
  • get_case
  • get_case_similar_observables
  • get_cases
  • get_cortex_analyzer
  • get_cortex_analyzer_job
  • get_observable
  • get_observable_analyzer_jobs
  • get_observables
  • get_task
  • get_tasks
  • import_alert_into_case
  • list_cortex_analyzers
  • list_cortex_analyzers_by_type
  • list_cortex_responders
  • merge_alert_into_case
  • merge_cases
  • promote_alert_to_case
  • run_observable_analyzer
  • run_observable_analyzers
  • share_observable
  • start_task
  • unfollow_alert
  • unshare_observable
  • update_alert
  • update_case
  • update_observable
  • update_task

Setup

Dependencies

This project depends on the official TheHive client library (thehive4py). It will be automatically installed when you install this package.

Environment Variables

Set the following environment variables:

HIVE_URL=<your-thehive-url>              # Required: TheHive instance URL (e.g., https://thehive.company.com:9000)
HIVE_API_KEY=<your-thehive-api-key>      # Required: API key for authenticating with TheHive instance

Usage with Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "thehive-mcp-server": {
      "command": "uvx",
      "args": ["thehive-mcp-server"],
      "env": {
        "HIVE_URL": "https://your-thehive-host:9000",
        "HIVE_API_KEY": "your-api-key"
      }
    }
  }
}

Alternative configuration using uv:

{
  "mcpServers": {
    "thehive-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp",
        "run",
        "python",
        "-m",
        "thehive_mcp.main"
      ],
      "env": {
        "HIVE_URL": "https://your-thehive-host:9000",
        "HIVE_API_KEY": "your-api-key"
      }
    }
  }
}

Replace /path/to/mcp with the actual path to the root directory where you've cloned the repository.

Manual Execution

You can also run the server manually:

# Install dependencies
pip install .

# Set environment variables
export HIVE_URL="http://localhost:9000"
export HIVE_API_KEY="your-api-key"

# Run the server
python -m thehive_mcp

FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.