How to add BloodHound-MCP-AI to Claude Desktop

Analyze BloodHound Active Directory attack paths with natural language via 75+ Cypher-backed MCP tools. Paste the config into ~/Library/Application Support/Claude/claude_desktop_config.json and restart Claude Desktop.

Last updated June 14, 2026 · 363★ · stdio · no auth

Claude Desktop config for BloodHound-MCP-AI

git clone https://github.com/MorDavid/BloodHound-MCP-AI.git && pip install -r requirements.txt

{
  "mcpServers": {
    "bloodhound-mcp-ai": {
      "command": "python",
      "args": [
        "<Your_Path>\\BloodHound-MCP.py"
      ],
      "env": {
        "BLOODHOUND_URI": "bolt://localhost:7687",
        "BLOODHOUND_USERNAME": "neo4j",
        "BLOODHOUND_PASSWORD": "bloodhoundcommunityedition"
      }
    }
  }
}

Setup steps

1Open Claude Desktop → Settings → Developer → Edit Config (this opens ~/Library/Application Support/Claude/claude_desktop_config.json).
2Paste the BloodHound-MCP-AI config below under the top-level "mcpServers" key.
3Fill in any placeholder secrets (API keys, paths) in the snippet.
4Save the file, then fully quit and reopen Claude Desktop.
5Open a chat and confirm BloodHound-MCP-AI's tools appear under the 🔌 tools menu.

Before you start

BloodHound 4.x+ with data collected from an Active Directory environment
Neo4j database with BloodHound data loaded
Python 3.8 or higher
An MCP client (e.g. Claude Desktop, Cursor)

What BloodHound-MCP-AI can do in Claude Desktop

Domain structure mapping

Query and map the structure of Active Directory domains, users, groups, computers, and their relationships.

Privilege escalation paths

Discover attack paths to high-value targets such as Domain Admins, including where Domain Users have local admin rights.

Kerberos security analysis

Identify Kerberos issues including Kerberoastable users and AS-REP Roasting candidates and their paths to privileged accounts.

Certificate services (ADCS) analysis

Map Active Directory Certificate Services vulnerabilities and abuse opportunities.

Active Directory hygiene assessment

Assess AD hygiene, e.g. finding inactive privileged accounts and other posture weaknesses.

NTLM relay attack analysis

Identify domain controllers and other targets vulnerable to NTLM relay attacks.

Delegation abuse analysis

Discover delegation abuse opportunities (e.g. unconstrained/constrained/resource-based delegation).

Security report generation

Generate comprehensive natural-language security reports about a domain for stakeholders.

Security

Connects to a Neo4j database holding BloodHound Active Directory data; configured with Neo4j credentials passed via env (BLOODHOUND_URI, BLOODHOUND_USERNAME, BLOODHOUND_PASSWORD). This is an offensive-security analysis tool: only run it against environments you are explicitly authorized to assess, treat BloodHound data as sensitive, and follow responsible-disclosure practices for any vulnerabilities found.

BloodHound-MCP-AI + Claude Desktop FAQ

Where is the Claude Desktop config file?

Claude Desktop reads MCP servers from ~/Library/Application Support/Claude/claude_desktop_config.json. Paste the BloodHound-MCP-AI config there under the "mcpServers" key and restart the client.

Is BloodHound-MCP-AI safe to use with Claude Desktop?

What does this server connect to?

It connects to the Neo4j graph database that BloodHound uses to store Active Directory relationship data, over the Bolt protocol. You provide the URI, username, and password via environment variables.

Does it need an API key?

No. Authentication is to your local Neo4j/BloodHound database via Neo4j credentials (env vars), not an external API key.

How many tools does it provide?

Over 75 specialized tools based on the original BloodHound CE Cypher queries, spanning domain mapping, privilege escalation, Kerberos, certificate services, AD hygiene, NTLM relay, and delegation abuse.

View repo Full BloodHound-MCP-AI page