graph-security-dashboard
Security Graph MCP Server
README
🛡️ Autonomous Security Graph Dashboard
An enterprise-grade, high-performance security graph visualization and investigation dashboard powered by FastAPI, Kùzu Graph Database, Cytoscape.js, and an embedded Model Context Protocol (MCP) Server.
🌟 Key Features
- Live Cloud Asset Discovery: Seamless integration with Azure Resource Graph (ARG) for instant discovery of Virtual Machines, Subnets, VNets, Storage Accounts, and Network Interfaces.
- Entra ID Identity Ingestion: High-priority extraction and paging of Entra ID user accounts (Global Admins, CISOs, SOC Analysts, Managers, active VM sessions) mapped to endpoint workloads.
- Single-Threaded Kùzu Graph Engine: Dedicated Python worker-thread queue architecture ensuring zero C++ lock contention or multithreading deadlocks.
- Model Context Protocol (MCP) Server: Native FastMCP integration (
src/backend/mcp_server.py) enabling AI agents (e.g., Claude, Antigravity) to query blast radiuses and inspect graph schemas via Cypher. - Interactive Cytoscape.js GUI: Dynamic canvas featuring Obsidian Dark aesthetic (
#111111canvas,#e93d82highlights), zoom controls, node search, layout transitions, and detailed inspector drawers. - Security Incident Correlation: Automatic mapping of live Defender Alerts & Sentinel Security Incidents to compromised cloud infrastructure.
🏗️ Architecture Overview
graph TD
A[Azure Tenant / ARG API] -->|Metadata Query| B[Scanner Engine src/backend/scanner.py]
C[Entra ID / Intune] -->|az ad user list| B
B -->|Ingest Nodes & Edges| D[Kùzu DB Queue Worker src/backend/db.py]
D -->|Cypher Queries| E[(Kùzu Embedded Graph DB)]
F[FastAPI Server src/backend/app.py] -->|GET /api/graph| D
F -->|POST /api/scan| B
G[React + Cytoscape UI src/frontend] <-->|SSE Logs / REST| F
H[AI Agents / MCP Clients] <-->|FastMCP src/backend/mcp_server.py| D
🚀 Quick Start Guide
Prerequisites
- Python 3.11+
- Node.js 18+ &
npm - Azure CLI (
az) (Optional for live tenant discovery)
1. Local Development Setup
Backend Setup
# Clone repository
git clone https://github.com/IamVigneshk/graph-security-dashboard.git
cd graph-security-dashboard
# Create virtual environment
python3 -m venv venv
source venv/bin/activate
# Install dependencies
pip install -r requirements.txt
# Start FastAPI backend server (Runs on http://localhost:8000)
PYTHONPATH=. python src/backend/app.py
Frontend Setup
# Open a new terminal and navigate to frontend directory
cd src/frontend
# Install dependencies
npm install
# Start Vite development server (Runs on http://localhost:5173)
npm run dev
2. Running with Docker Compose
To launch both the backend API server and frontend build in isolated containers:
docker-compose up --build
Access the application at http://localhost:8000.
🔌 Model Context Protocol (MCP) Server Integration
The backend includes a standalone FastMCP server allowing LLMs and AI agents to query the security graph safely:
Running the MCP Server
PYTHONPATH=. python src/backend/mcp_server.py
Exposed MCP Tools
get_graph_schema(): Returns graph node types (Machine,CloudResource,Incident,Alert,User) and relationship schemas (:AFFECTS,:HOSTED_IN,:LOGGED_IN_TO,:INCLUDES).read_cypher_query(query): Executes read-only Cypher queries against the embedded Kùzu database. Write operations (CREATE,DELETE,SET) are strictly blocked.
Example MCP Cypher Query
MATCH (i:Incident)-[:AFFECTS]->(m:Machine)-[:HOSTED_IN]->(r:CloudResource)
RETURN i.title, m.name, r.type
📡 API Endpoints Reference
| Method | Endpoint | Description |
|---|---|---|
GET |
/api/health |
Healthcheck returning Kùzu engine status |
POST |
/api/scan |
Triggers Azure & Defender scan via SSE log stream |
GET |
/api/graph |
ReturnsCytoscape-formatted graph nodes & edges |
🎨 UI Aesthetic & Design System
- Background:
#111111(Obsidian Dark) - Cards & Drawers:
#191919/#1a1a1awith glassmorphic borders - Primary Accent:
#e93d82(Magenta Glow) - Severity Colors: Critical (
#ff3366), High (#ff6600), Medium (#ffcc00), Low (#00ccff)
📄 License
MIT License. Developed by Vignesh K.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
E2B
Using MCP to run code via e2b.