Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT and other tools is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AI Fire in da houseCheck it out free

roadrecon_mcp_server

MCP.Pizza Chef: atomicchonk

The roadrecon_mcp_server is an MCP server that integrates with the ROADRecon Azure AD security toolset, providing AI assistants like Claude with structured access to Azure AD data for security analysis. It offers resources to fetch data, tools to run security checks, and pre-built prompts for common security tasks, facilitating automated and intelligent security assessments within Azure AD environments.

Use This MCP server To

Access Azure AD data from ROADRecon for AI-driven analysis Run automated security assessments on Azure AD configurations Use pre-built prompts to perform common security tasks Integrate ROADRecon data into AI workflows for threat detection Enable AI assistants to query ROADRecon data in real time Automate security report generation from ROADRecon insights

README

ROADrecon MCP Server

This MCP (Model Context Protocol) server provides AI assistants like Claude with access to your ROADRecon Azure AD data for security analysis.

The amazing ROADtools suite by dirkjanm can be found here: ROADRecon

Features

  • Resources: Access Azure AD data from your ROADRecon instance
  • Tools: Run security analysis on the data
  • Prompts: Pre-built analysis templates for common security tasks

Prerequisites

  • Python 3.8+
  • A running ROADRecon instance with the web GUI accessible
  • MCP-compatible client (Claude Desktop, etc.)

Installation

  1. Clone this repository
  2. Install dependencies:
    pip install -r requirements.txt
    

Usage

Running the server

  1. Make sure your ROADRecon GUI is running (default: http://localhost:5000)

  2. Run the MCP server:

    python roadrecon_mcp_server.py
    
  3. To specify a different ROADRecon URL:

    ROADRECON_URL=http://localhost:8080 python roadrecon_mcp_server.py
    

Connecting with Claude Desktop

  1. Open Claude Desktop
  2. Go to Settings → Servers → Add Server
  3. Select "Add from running server"
  4. The server should appear in the list - click "Install"

More details on this step can be found here: https://modelcontextprotocol.io/quickstart/server

Using in Claude

Once connected, Claude can:

  • Access Azure AD data via resources (e.g., roadrecon://users)
  • Run security analysis with tools (e.g., find_privileged_users)
  • Use pre-built prompts for common security tasks

Example Queries

  • "Analyze the MFA status of users in this Azure AD tenant"
  • "Find all users with privileged roles"
  • "Check for applications with secrets or certificates"
  • "Analyze the overall security posture of this Azure AD environment"
walkthrough_video.mp4

Resources Available

  • roadrecon://stats - Summary statistics
  • roadrecon://users - All users
  • roadrecon://users/{id} - User details
  • roadrecon://groups - All groups
  • roadrecon://groups/{id} - Group details
  • roadrecon://applications - All applications
  • roadrecon://applications/{id} - Application details
  • roadrecon://serviceprincipals - All service principals
  • roadrecon://serviceprincipals/{id} - Service principal details
  • roadrecon://devices - All devices
  • roadrecon://mfa - MFA status for all users
  • roadrecon://directoryroles - All directory roles
  • roadrecon://roledefinitions - All role definitions
  • roadrecon://approles - All app role assignments
  • roadrecon://oauth2permissions - All OAuth2 permission grants
  • roadrecon://tenantdetails - Tenant details

Tools Available

  • find_privileged_users() - Find users with high-privilege roles
  • analyze_mfa_status() - Analyze MFA deployment across users
  • find_applications_with_secrets() - Find applications with secrets/certificates
  • analyze_groups() - Analyze group types and membership
  • identify_stale_accounts() - Find accounts that haven't logged in or changed password within a specified period
  • analyze_pim_implementation() - Assess Privileged Identity Management implementation
  • analyze_service_principal_credentials() - Find over-permissioned service principals with long-lived credentials
  • analyze_legacy_authentication() - Identify risks from legacy authentication protocols that bypass MFA
  • analyze_conditional_access_policies(file_path: str = "") - Analyze conditional access policies from an HTML file. Looks for the file at "C:\Temp\caps.html" by default, or prompts the user to specify a file path if not found.

Prompts Available

  • analyze_security_posture - Comprehensive security analysis
  • analyze_privileged_access - Analysis of privileged access model
  • investigate_application_risks - Application security risk assessment
  • analyze_identity_security - Identity security configuration analysis
  • analyze_stale_accounts - Analysis of inactive user accounts
  • analyze_privileged_access_management - PIM implementation assessment
  • analyze_service_principal_security - Service principal credential risk analysis
  • analyze_legacy_authentication_risks - Legacy authentication protocol risk assessment
  • analyze_conditional_access - Analysis of conditional access policies and recommendations
  • comprehensive_security_review - Complete security review of the entire environment

License

MIT

roadrecon_mcp_server FAQ

How do I install the roadrecon_mcp_server?
Clone the repository and run 'pip install -r requirements.txt' to install dependencies.
What prerequisites are needed to run this MCP server?
You need Python 3.8+, a running ROADRecon instance with web GUI, and an MCP-compatible client like Claude Desktop.
How do I start the roadrecon_mcp_server?
Ensure ROADRecon GUI is running, then execute 'python roadrecon_mcp_server.py' to start the server.
Can I connect the server to a ROADRecon instance on a different URL?
Yes, specify the URL using the ROADRECON_URL environment variable before running the server.
What AI assistants are compatible with this MCP server?
It supports AI assistants compatible with MCP such as Claude, OpenAI GPT models, and Anthropic's Claude.
What kind of data does this MCP server expose?
It exposes Azure AD security data collected by ROADRecon for analysis and querying.
Are there pre-built prompts available?
Yes, the server includes pre-built analysis templates for common security tasks to streamline workflows.
Is this MCP server limited to Azure AD data?
Yes, it specifically integrates with ROADRecon to provide Azure AD security data for analysis.