1. Conduid
  2. Developer Tools
  3. Graphsecurityincidents
MCP server · Developer Tools

Graphsecurityincidents

This project is a sophisticated **Model Context Protocol (MCP) server** for Microsoft Graph Security API Incident and Alert management

Unclaimed MIT last commit a year ago devtools
54Fair

Scored 4 months ago · breakdown

About Graphsecurityincidents

Graphsecurityincidents is an MCP server published by LaurieRhodes in the Developer Tools category: this project is a sophisticated **Model Context Protocol (MCP) server** for Microsoft Graph Security API Incident and Alert management. It has been installed 0 times through Conduid.

The repository has 15 stars and 2 forks, with the last commit a year ago. Six months or more without a commit doesn't mean the server is broken, but check the open issues (0) before depending on it in production.

Install

Install
npx mcp-graphsecurityincidents

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

Ask AI about Graphsecurityincidents

Powered by Claude · Grounded in docs

I know everything about Graphsecurityincidents. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

  • ·README presentNot checked yet.
  • ·License declaredNot checked yet.
  • ·Tests presentNot checked yet.
  • ·Dependencies pinnedNot checked yet.
  • ·No dynamic code executionNot checked yet.
  • !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.

README

Microsoft Graph Security MCP Server

MIT License Go Version

A Model Context Protocol (MCP) server implementation for Microsoft Graph Security APIs, providing intelligent field selection and token optimisation for AI assistant interactions.

Overview

This server implements the Model Context Protocol to expose Microsoft Graph Security APIs (alerts, incidents) to MCP-compatible AI assistants. It includes intelligent field selection using native Graph API $select parameters to reduce response sizes and token usage.

Key Features

  • Universal Tools: 10 tools that work consistently across security entity types
  • Field Selection: Automatic optimization using Microsoft Graph $select parameters
  • Intent Detection: Query analysis to determine appropriate response detail level
  • TTL Context Caching: Reduces redundant context provision
  • Authentication: Support for Azure AD app credentials and managed identity
  • Thread Safety: Concurrent request handling with proper synchronization

Architecture

┌─────────────────┐    ┌──────────────┐    ┌─────────────────┐    ┌─────────────────┐
│ MCP Client      │    │ MCP Server   │    │ Universal       │    │ Microsoft       │
│ (AI Assistant)  │◄──►│ (This Tool)  │◄──►│ Graph Client    │◄──►│ Graph Security  │
└─────────────────┘    └──────────────┘    └─────────────────┘    └─────────────────┘

Core Components

  • MCP Server (internal/mcp): Handles JSON-RPC 2.0 protocol communication
  • Universal Client (internal/graph/framework): Graph API abstraction with field selection
  • Entity System (internal/entities): Schema definitions with field priority metadata
  • Context Management (internal/context): TTL-based context optimization
  • Authentication (internal/auth): Azure AD credential handling

Installation

Prerequisites

  • Go 1.21 or later
  • Azure AD application with Graph Security permissions
  • Microsoft Graph Security data (Defender for Endpoint, etc.)

Build from Source

git clone https://github.com/LaurieRhodes/MCP-GraphSecurityIncidents.git
cd MCP-GraphSecurityIncidents
go build -o graph-security-incidents.exe ./cmd/server

Configuration

Required Azure Permissions

Grant these Microsoft Graph permissions to your Azure AD application:

SecurityAlert.Read.All
SecurityAlert.ReadWrite.All  
SecurityIncident.Read.All
SecurityIncident.ReadWrite.All

Configuration File

Create config.json:

{
  "auth": {
    "type": "app",
    "tenant": "your-tenant-id",
    "clientId": "your-client-id", 
    "clientSecret": "your-client-secret"
  },
  "graph": {
    "apiVersion": "beta",
    "endpoint": "https://graph.microsoft.com",
    "requestTimeout": 30,
    "security": {
      "alertsApiVersion": "v2",
      "incidentsApiVersion": "v1"
    }
  },
  "server": {
    "name": "mcp-graph-security-incidents",
    "version": "1.0.0"
  }
}

MCP Client Configuration

Add to your MCP client configuration:

{
  "mcpServers": {
    "graph-security": {
      "command": "/path/to/graph-security-incidents",
      "args": []
    }
  }
}

Note that Claude is a poor LLM for management of Security data due to its safety protocols preventing sensitive data to be divulged. LLM Testing Results: Provides detailed analysis of current LLM with tool using capabilities against this MCP server.

Available Tools

Universal Entity Tools

Tool Description Parameters
graph_entity_list List security entities entityType, filter, top, orderBy, intent
graph_entity_get Get specific entity entityType, entityId, expand
graph_entity_update Update entity properties entityType, entityId, properties
graph_entity_comment Add comment to entity entityType, entityId, comment
graph_entity_navigate Navigate entity relationships sourceEntityType, sourceEntityId, targetEntityType
graph_entity_list_next Get next page of results nextLink
graph_entity_schema Get entity schema info entityType, operation

Context Management Tools

Tool Description Parameters
context_discover Discover entity capabilities entityType, focusArea
context_stats Get context usage statistics None
context_configure Configure context behavior action, level, seconds, toolName

Supported Entity Types

  • alert: Microsoft Graph Security alerts
  • incident: Microsoft Graph Security incidents

Field Selection System

Query Intents

The server automatically detects query intent and selects appropriate fields:

type QueryIntent string

const (
    IntentOverview  QueryIntent = "overview"  // Essential fields only
    IntentStandard  QueryIntent = "standard"  // Essential + operational fields  
    IntentComplete  QueryIntent = "complete"  // All fields
)

Field Priorities

Fields are categorized by priority for selection optimization:

  • Essential: Core identification fields (id, displayName, severity, status)
  • Standard: Operational fields (dates, assignment, classification)
  • Complete: All remaining fields (descriptions, evidence, comments)

Context Management

TTL-Based Optimization

Context is provided based on time-to-live (TTL) settings:

  • Default TTL: 1 hour
  • Context Levels: none, minimal, standard, complete
  • Per-Tool Tracking: Independent TTL for each tool

## Development

Project Structure

├── cmd/server/           # Main server entry point
├── internal/
│   ├── auth/            # Authentication implementations
│   ├── config/          # Configuration management
│   ├── context/         # Smart context management
│   ├── entities/        # Entity system and schemas
│   ├── graph/           # Microsoft Graph client
│   ├── mcp/             # MCP protocol implementation
│   ├── schemas/         # Legacy schema definitions
│   ├── tools/           # Tool registration and schemas
│   └── utils/           # Utility functions
├── docs/                # Documentation
└── config.example.json  # Example configuration

Adding New Entity Types

  1. Create entity package in internal/entities/
  2. Implement Entity interface
  3. Define schema with field priorities
  4. Register via init() function
  5. Add context providers and validators

Building

go build -o graph-security-incidents ./cmd/server

Documentation

Contributing

This is a personal development and I doubt there is much interest in extending this framework to a full community development project.

License

This project is licensed under the MIT License - see LICENSE file for details.

Related Projects

README mirrored from the source repository 4 months ago. The original is authoritative.

Questions

About Graphsecurityincidents

How do I install Graphsecurityincidents?

Run npx mcp-graphsecurityincidents, then add the server to your MCP client's configuration. Conduid has recorded 0 installs, so the command is known to work with current clients.

Is Graphsecurityincidents safe to use with an AI agent?

Its trust score is 54 out of 100 (fair). It passes 0 of 1 static security checks; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Graphsecurityincidents still maintained?

The last commit was a year ago, with 0 open issues. That's long enough that you should check whether the maintainer is responding to issues before depending on it.